Using rocprofv3

Using rocprofv3#

rocprofv3 is a CLI tool that helps you quickly optimize applications and understand the low-level kernel details without requiring any modification in the source code. It’s backward compatible with its predecessor, rocprof, and provides more features for application profiling with better accuracy.

The following sections demonstrate the use of rocprofv3 for application tracing and kernel profiling using various command-line options.

rocprofv3 is installed with ROCm under /opt/rocm/bin. To use the tool from anywhere in the system, export PATH variable:

export PATH=$PATH:/opt/rocm/bin

Before you start tracing or profiling your HIP application using rocprofv3, build the application using:

cmake -B <build-directory> <source-directory> -DCMAKE_PREFIX_PATH=/opt/rocm
cmake --build <build-directory> --target all --parallel <N>

Options#

Here is the sample of commonly used rocprofv3 command-line options. Some options are used for application tracing and some for kernel profiling while the output control options control the presentation and redirection of the generated output.

Table 1 rocprofv3 options#
Option	Description	Use
`-i` \| `--input`	Specifies the input file. JSON and YAML formats support configuration of all command-line options whereas the text format only supports specifying HW counters.	Run Configuration
`-d` \| `--output-directory`	Specifies the path for the output files. Supports special keys: `%hostname%`, `%pid%`, `%rank%`, etc.	Output control
`-o` \| `--output-file`	Specifies the name of the output file. Note that this name is appended to the default names (_api_trace or counter_collection.csv) of the generated files’. Supports special keys: `%hostname%`, `%pid%`, `%rank%`, etc.	Output control
`--output-format`	For adding output format (supported formats: csv, json, pftrace)	Output control
`-r` \| `--runtime-trace`	Collects HIP (runtime), memory copy, marker, scratch memory, and kernel dispatch traces.	Application Tracing
`-s` \| `--sys-trace`	Collects HIP, HSA, memory copy, marker, scratch memory, and kernel dispatch traces.	Application Tracing
`--hip-trace`	Collects HIP runtime and compiler traces.	Application tracing
`--kernel-trace`	Collects kernel dispatch traces.	Application tracing
`--marker-trace`	Collects marker (ROC-TX) traces.	Application tracing
`--memory-copy-trace`	Collects memory copy traces.	Application tracing
`--scratch-memory-trace`	Collects scratch memory operations traces.	Application tracing
`--hsa-trace`	Collects HSA API traces.	Application tracing
`--hip-runtime-trace`	Collects HIP runtime API traces.	Application tracing
`--hsa-core-trace`	Collects HSA API traces (core API).	Application tracing
`--hsa-amd-trace`	Collects HSA API traces (AMD-extension API).	Application tracing
`--stats`	For Collecting statistics of enabled tracing types	Application tracing
`-p` \| `--summary`	Display summary of collected data	Application tracing
`--kernel-include-regex`	Include the kernels matching this filter.	Kernel Dispatch Counter Collection
`--kernel-exclude-regex`	Exclude the kernels matching this filter.	Kernel Dispatch Counter Collection
`--kernel-iteration-range`	Iteration range for each kernel that match the filter [start-stop].	Kernel Dispatch Counter Collection
`-L` \| `--list-metrics`	List metrics for counter collection.	Kernel Dispatch Counter Collection
`-M` \| `--mangled-kernels`	Overrides the default demangling of kernel names.	Output control
`-T` \| `--truncate-kernels`	Truncates the demangled kernel names for improved readability.	Output control
`--output-format`	For adding output format (supported formats: csv, json, pftrace, otf2)	Output control
`--preload`	Libraries to prepend to LD_PRELOAD (usually for sanitizers)	Extension
`--perfetto-backend {inprocess,system}`	Perfetto data collection backend. ‘system’ mode requires starting traced and perfetto daemons	Extension
`--perfetto-buffer-size KB`	Size of buffer for perfetto output in KB. default: 1 GB	Extension
`--perfetto-buffer-fill-policy {discard,ring_buffer}`	Policy for handling new records when perfetto has reached the buffer limit	Extension
`--perfetto-shmem-size-hint KB`	Perfetto shared memory size hint in KB. default: 64 KB	Extension

To see exhaustive list of rocprofv3 options, run:

rocprofv3 --help

Application tracing#

Application tracing provides the big picture of a program’s execution by collecting data on the execution times of API calls and GPU commands, such as kernel execution, async memory copy, and barrier packets. This information can be used as the first step in the profiling process to answer important questions, such as how much percentage of time was spent on memory copy and which kernel took the longest time to execute.

To use rocprofv3 for application tracing, run:

rocprofv3 <tracing_option> -- <app_relative_path>

HIP trace#

HIP trace comprises execution traces for the entire application at the HIP level. This includes HIP API functions and their asynchronous activities at the runtime level. In general, HIP APIs directly interact with the user program. It is easier to analyze HIP traces as you can directly map them to the program.

To trace HIP runtime APIs, use:

rocprofv3 --hip-trace -- < app_relative_path >

The above command generates a hip_api_trace.csv file prefixed with the process ID.

$ cat 238_hip_api_trace.csv

Here are the contents of hip_api_trace.csv file:

Table 2 HIP runtime api trace#
Domain	Function	Process_Id	Thread_Id	Correlation_Id	Start_Timestamp	End_Timestamp
HIP_COMPILER_API	__hipRegisterFatBinary	208	208	1	1508780270085955	1508780270096795
HIP_COMPILER_API	__hipRegisterFunction	208	208	2	1508780270104242	1508780270115355
HIP_COMPILER_API	__hipPushCallConfiguration	208	208	3	1508780613897816	1508780613898701
HIP_COMPILER_API	__hipPopCallConfiguration	208	208	4	1508780613901714	1508780613902200

To trace HIP compile time APIs, use:

rocprofv3 --hip-compiler-trace -- < app_relative_path >

The above command generates a hip_api_trace.csv file prefixed with the process ID.

$ cat 208_hip_api_trace.csv

Here are the contents of hip_api_trace.csv file:

Table 3 HIP compile time api trace#
Domain	Function	Process_Id	Thread_Id	Correlation_Id	Start_Timestamp	End_Timestamp
HIP_COMPILER_API	__hipRegisterFatBinary	208	208	1	1508780270085955	1508780270096795
HIP_COMPILER_API	__hipRegisterFunction	208	208	2	1508780270104242	1508780270115355
HIP_COMPILER_API	__hipPushCallConfiguration	208	208	3	1508780613897816	1508780613898701
HIP_COMPILER_API	__hipPopCallConfiguration	208	208	4	1508780613901714	1508780613902200

For the description of the fields in the output file, see Output file fields.

HSA trace#

The HIP runtime library is implemented with the low-level HSA runtime. HSA API tracing is more suited for advanced users who want to understand the application behavior at the lower level. In general, tracing at the HIP level is recommended for most users. You should use HSA trace only if you are familiar with HSA runtime.

HSA trace contains the start and end time of HSA runtime API calls and their asynchronous activities.

rocprofv3 --hsa-trace -- < app_relative_path >

The above command generates a hsa_api_trace.csv file prefixed with process ID. Note that the contents of this file have been truncated for demonstration purposes.

$ cat 197_hsa_api_trace.csv

Here are the contents of hsa_api_trace.csv file:

Table 4 HSA api trace#
Domain	Function	Process_Id	Thread_Id	Correlation_Id	Start_Timestamp	End_Timestamp
HSA_CORE_API	hsa_system_get_major_extension_table	197	197	1	1507843974724237	1507843974724947
HSA_CORE_API	hsa_agent_get_info	197	197	3	1507843974754471	1507843974755014
HSA_AMD_EXT_API	hsa_amd_memory_pool_get_info	197	197	5	1507843974761705	1507843974762398
HSA_AMD_EXT_API	hsa_amd_memory_pool_get_info	197	197	6	1507843974763901	1507843974764030
HSA_AMD_EXT_API	hsa_amd_memory_pool_get_info	197	197	7	1507843974765121	1507843974765224
HSA_AMD_EXT_API	hsa_amd_memory_pool_get_info	197	197	8	1507843974766196	1507843974766328
HSA_AMD_EXT_API	hsa_amd_memory_pool_get_info	197	197	9	1507843974767534	1507843974767641
HSA_AMD_EXT_API	hsa_amd_memory_pool_get_info	197	197	10	1507843974768639	1507843974768779
HSA_AMD_EXT_API	hsa_amd_agent_iterate_memory_pools	197	197	4	1507843974758768	1507843974769238
HSA_CORE_API	hsa_agent_get_info	197	197	11	1507843974771091	1507843974771537

For the description of the fields in the output file, see Output file fields.

Marker trace#

In certain situations, such as debugging performance issues in large-scale GPU programs, API-level tracing might be too fine-grained to provide a big picture of the program execution. In such cases, it is helpful to define specific tasks to be traced.

To specify the tasks for tracing, enclose the respective source code with the API calls provided by the ROCTx library. This process is also known as instrumentation. As the scope of code for instrumentation is defined using the enclosing API calls, it is called a range. A range is a programmer-defined task that has a well-defined start and end code scope. You can also refine the scope specified within a range using further nested ranges. rocprofv3 also reports the timelines for these nested ranges.

Here is a list of useful APIs for code instrumentation.

roctxMark: Inserts a marker in the code with a message. Creating marks help you see when a line of code is executed.
roctxRangeStart: Starts a range. Different threads can start ranges.
roctxRangePush: Starts a new nested range.
roctxRangePop: Stops the current nested range.
roctxRangeStop: Stops the given range.

Note

To use rocprofv3 for marker tracing, including and linking to old ROCTx works but it is recommended to switch to new ROCTx because it has been extended with new APIs. To use new ROCTx, please include header "rocprofiler-sdk-roctx/roctx.h" and link your application with librocprofiler-sdk-roctx.so. Above list of APIs is not exhaustive. See public header file "rocprofiler-sdk-roctx/roctx.h" for full list.

See how to use ROCTx APIs in the MatrixTranspose application below:

#include <rocprofiler-sdk-roctx/roctx.h>

roctxMark("before hipLaunchKernel");
int rangeId = roctxRangeStart("hipLaunchKernel range");
roctxRangePush("hipLaunchKernel");

// Launching kernel from host
hipLaunchKernelGGL(matrixTranspose, dim3(WIDTH/THREADS_PER_BLOCK_X, WIDTH/THREADS_PER_BLOCK_Y), dim3(THREADS_PER_BLOCK_X, THREADS_PER_BLOCK_Y), 0,0,gpuTransposeMatrix,gpuMatrix, WIDTH);

roctxMark("after hipLaunchKernel");

// Memory transfer from device to host
roctxRangePush("hipMemcpy");

hipMemcpy(TransposeMatrix, gpuTransposeMatrix, NUM * sizeof(float), hipMemcpyDeviceToHost);

roctxRangePop();  // for "hipMemcpy"
roctxRangePop();  // for "hipLaunchKernel"
roctxRangeStop(rangeId);

To trace the API calls enclosed within the range, use:

rocprofv3 --marker-trace -- < app_relative_path >

Running the preceding command generates a marker_api_trace.csv file prefixed with the process ID.

$ cat 210_marker_api_trace.csv

Here are the contents of marker_api_trace.csv file:

Table 5 Marker api trace#
Domain	Function	Process_Id	Thread_Id	Correlation_Id	Start_Timestamp	End_Timestamp
MARKER_CORE_API	before hipLaunchKernel	717	717	1	1520113899312225	1520113899312225
MARKER_CORE_API	after hipLaunchKernel	717	717	4	1520113900128482	1520113900128482
MARKER_CORE_API	hipMemcpy	717	717	5	1520113900141100	1520113901483408
MARKER_CORE_API	hipLaunchKernel	717	717	3	1520113899684965	1520113901491622
MARKER_CORE_API	hipLaunchKernel range	717	0	2	1520113899682208	1520113901495882

For the description of the fields in the output file, see Output file fields.

Kernel trace#

To trace kernel dispatch traces, use:

rocprofv3 --kernel-trace -- < app_relative_path >

The above command generates a kernel_trace.csv file prefixed with the process ID.

$ cat 199_kernel_trace.csv

Here are the contents of kernel_trace.csv file:

Table 6 Kernel trace#
Kind	Agent_Id	Queue_Id	Thread_Id	Dispatch_Id	Kernel_Id	Kernel_Name	Correlation_Id	Start_Timestamp	End_Timestamp	Private_Segment_Size	Group_Segment_Size	Workgroup_Size_X	Workgroup_Size_Y	Workgroup_Size_Z	Grid_Size_X	Grid_Size_Y	Grid_Size_Z
KERNEL_DISPATCH	1	1	69	1	16	void addition_kernel<float>(float, float const, float const*, int, int)	1451	8819330200067564	8819330200116308	0	0	64	1	1	1024	1024	1
KERNEL_DISPATCH	1	2	69	5	16	void addition_kernel<float>(float, float const, float const*, int, int)	1484	8819330200118678	8819330200219573	0	0	64	1	1	1024	1024	1
KERNEL_DISPATCH	1	1	69	2	19	subtract_kernel(float, float const, float const*, int, int)	1459	8819330200120456	8819330200223721	0	0	64	1	1	1024	1024	1
KERNEL_DISPATCH	1	3	69	9	16	void addition_kernel<float>(float, float const, float const*, int, int)	1517	8819330200152902	8819330200283428	0	0	64	1	1	1024	1024	1
KERNEL_DISPATCH	1	4	69	13	16	void addition_kernel<float>(float, float const, float const*, int, int)	1550	8819330200187127	8819330200320468	0	0	64	1	1	1024	1024	1
KERNEL_DISPATCH	1	2	69	6	19	subtract_kernel(float, float const, float const*, int, int)	1492	8819330200225499	8819330200364618	0	0	64	1	1	1024	1024	1
KERNEL_DISPATCH	1	1	69	3	18	multiply_kernel(float, float const, float const*, int, int)	1467	8819330200229796	8819330200369359	0	0	64	1	1	1024	1024	1

For the description of the fields in the output file, see Output file fields.

Memory copy trace#

To trace memory moves across the application, use:

rocprofv3 –-memory-copy-trace -- < app_relative_path >

The above command generates a memory_copy_trace.csv file prefixed with the process ID.

$ cat 197_memory_copy_trace.csv

Here are the contents of memory_copy_trace.csv file:

Table 7 Memory copy trace#
Kind	Direction	Source_Agent_Id	Destination_Agent_Id	Correlation_Id	Start_Timestamp	End_Timestamp
MEMORY_COPY	MEMORY_COPY_HOST_TO_DEVICE	0	1	0	14955949675563	14955950239443
MEMORY_COPY	MEMORY_COPY_DEVICE_TO_HOST	1	0	0	14955952733485	14955953315285

For the description of the fields in the output file, see Output file fields.

Runtime trace#

This is a short-hand option which attempts to target the most relevant tracing options for a standard user by excluding tracing the HSA runtime API and HIP compiler API.

The HSA runtime API is excluded because it is a lower-level API upon which HIP and OpenMP target are built and thus, tends to be an implementation detail not relevant to most users. The HIP compiler API is excluded because these are functions which are automatically inserted during HIP compilation and thus, also tend to be implementation details which are not relevant to most users.

At present, –runtime-trace enables tracing the HIP runtime API, the marker API, kernel dispatches, and memory operations (copies and scratch).

rocprofv3 –-runtime-trace -- < app_relative_path >

Running the above command generates hip_api_trace.csv, kernel_trace.csv, memory_copy_trace.csv, scratch_memory_trace.csv,and marker_api_trace.csv (if ROCTx APIs are specified in the application) files prefixed with the process ID.

System trace#

This is an all-inclusive option to collect all the above-mentioned traces.

rocprofv3 –-sys-trace -- < app_relative_path >

Running the above command generates hip_api_trace.csv, hsa_api_trace.csv, kernel_trace.csv, memory_copy_trace.csv, and marker_api_trace.csv (if ROCTx APIs are specified in the application) files prefixed with the process ID.

Scratch memory trace#

This option collects scratch memory operation’s traces. Scratch is an address space on AMD GPUs, which is roughly equivalent to the local memory in NVIDIA CUDA. The local memory in CUDA is a thread-local global memory with interleaved addressing, which is used for register spills or stack space. With this option, you can trace when the rocr runtime allocates, frees, and tries to reclaim scratch memory.

rocprofv3 --scratch-memory-trace -- < app_relative_path >

Stats#

This option collects statistics for the enabled tracing types. For example, to collect statistics of HIP APIs, when HIP trace is enabled. A higher percentage in statistics can help user focus on the API/function that has taken the most time:

rocprofv3 --stats --hip-trace  -- < app_relative_path >

The above command generates a hip_api_stats.csv and hip_api_trace file prefixed with the process ID.

$ cat hip_api_stats.csv

Here are the contents of hip_api_stats.csv file:

Table 8 HIP stats#
Name	Calls	TotalDurationNs	AverageNs	Percentage	MinNs	MaxNs	StdDev
hipStreamCreateWithFlags	4	262497406	65624351.500000	85.15	3991286	249121840	122332531.343496
hipGetDeviceCount	1	32505687	32505687.000000	10.54	32505687	32505687	0.00000000e+00
hipHostMalloc	12	6096409	508034.083333	1.98	443793	548024	39236.753678
hipFree	12	1994421	166201.750000	0.6470	7790	1036046	299086.860470
hipMemcpyAsync	12	1368378	114031.500000	0.4439	2490	764044	249308.051619
hipMallocAsync	12	927255	77271.250000	0.3008	51540	107671	20487.475966
hipStreamSynchronize	12	870486	72540.500000	0.2824	140	866606	250065.900069
hipLaunchKernel	16	692734	43295.875000	0.2247	1000	670044	167133.656647
hipStreamDestroy	4	619905	154976.250000	0.2011	92901	339252	122852.320356
hipDeviceSynchronize	4	404252	101063.000000	0.1311	570	385212	189518.505401
hipHostFree	12	271202	22600.166667	0.0880	11950	34950	7480.268600
__hipRegisterFatBinary	1	9000	9000.000000	2.920e-03	9000	9000	0.00000000e+00
__hipRegisterFunction	4	6150	1537.500000	1.995e-03	230	5370	2555.091323
__hipPushCallConfiguration	16	2460	153.750000	7.980e-04	70	1140	267.503894
__hipPopCallConfiguration	16	2000	125.000000	6.488e-04	70	680	151.613544
hipGetLastError	16	1270	79.375000	4.120e-04	50	440	96.295985
hipSetDevice	1	660	660.000000	2.141e-04	660	660	0.00000000e+00

For the description of the fields in the output file, see Output file fields.

Kernel profiling#

The application tracing functionality allows you to evaluate the duration of kernel execution but is of little help in providing insight into kernel execution details. The kernel profiling functionality allows you to select kernels for profiling and choose the basic counters or derived metrics to be collected for each kernel execution, thus providing a greater insight into kernel execution.

For a comprehensive list of counters available on MI200, see MI200 performance counters and metrics.

Input file#

To collect the desired basic counters or derived metrics or tracing, mention them in an input file. The input file could be in text (.txt), yaml (.yaml/.yml), or JSON (.json) format.

In the input text file, the line consisting of the counter or metric names must begin with pmc. The number of basic counters or derived metrics that can be collected in one run of profiling are limited by the GPU hardware resources. If too many counters or metrics are selected, the kernels need to be executed multiple times to collect them. For multi-pass execution, include multiple pmc rows in the input file. Counters or metrics in each pmc row can be collected in each application run.

The JSON and YAML files supports all the command line options and it can be used to configure both tracing and profiling. The input file has an array of profiling/tracing configurations called jobs. Each job is used to configure profiling/tracing for an application execution. The input schema of these files is given below.

Properties#

``jobs`` (array): rocprofv3 input data per application run.
- Items (object): data for rocprofv3.
  - ``pmc`` (array): list of counters to collect.
  - ``kernel_include_regex`` (string): Include the kernels matching this filter.
  - ``kernel_exclude_regex`` (string): Exclude the kernels matching this filter.
  - ``kernel_iteration_range`` (string): Iteration range for each kernel that match the filter [start-stop].
  - ``hip_trace`` (boolean): For Collecting HIP Traces (runtime + compiler).
  - ``hip_runtime_trace`` (boolean): For Collecting HIP Runtime API Traces.
  - ``hip_compiler_trace`` (boolean): For Collecting HIP Compiler generated code Traces.
  - ``marker_trace`` (boolean): For Collecting Marker (ROCTx) Traces.
  - ``kernel_trace`` (boolean): For Collecting Kernel Dispatch Traces.
  - ``memory_copy_trace`` (boolean): For Collecting Memory Copy Traces.
  - ``scratch_memory_trace`` (boolean): For Collecting Scratch Memory operations Traces.
  - ``stats`` (boolean): For Collecting statistics of enabled tracing types.
  - ``hsa_trace`` (boolean): For Collecting HSA Traces (core + amd + image + finalizer).
  - ``hsa_core_trace`` (boolean): For Collecting HSA API Traces (core API).
  - ``hsa_amd_trace`` (boolean): For Collecting HSA API Traces (AMD-extension API).
  - ``hsa_finalize_trace`` (boolean): For Collecting HSA API Traces (Finalizer-extension API).
  - ``hsa_image_trace`` (boolean): For Collecting HSA API Traces (Image-extension API).
  - ``sys_trace`` (boolean): For Collecting HIP, HSA, Marker (ROCTx), Memory copy, Scratch memory, and Kernel dispatch traces.
  - ``mangled_kernels`` (boolean): Do not demangle the kernel names.
  - ``truncate_kernels`` (boolean): Truncate the demangled kernel names.
  - ``output_file`` (string): For the output file name.
  - ``output_directory`` (string): For adding output path where the output files will be saved.
  - ``output_format`` (array): For adding output format (supported formats: csv, json, pftrace, otf2).
  - ``list_metrics`` (boolean): List the metrics.
  - ``log_level`` (string): fatal, error, warning, info, trace.
  - ``preload`` (array): Libraries to prepend to LD_PRELOAD (usually for sanitizers).

$ cat input.txt

pmc: GPUBusy SQ_WAVES
pmc: GRBM_GUI_ACTIVE

$ cat input.json

{
    "jobs": [
    {
        "pmc": ["SQ_WAVES", "GRBM_COUNT", "GUI_ACTIVE"]
    },
    {
        "pmc": ["FETCH_SIZE", "WRITE_SIZE"],
        "kernel_include_regex": ".*_kernel",
        "kernel_exclude_regex": "multiply",
        "kernel_iteration_range": "[1-2]","[3-4]"
        "output_file": "out",
        "output_format": [
                "csv",
                "json"
        ],
        "truncate_kernels": true
    ]
}

  $ cat input.yaml

jobs:
  - pmc:
      - SQ_WAVES
      - GRBM_COUNT
      - GUI_ACTIVE
      - 'TCC_HIT[1]'
      - 'TCC_HIT[2]'
  - pmc:
      - FETCH_SIZE
      - WRITE_SIZE

Command-line#

Desired counters can now be collected as command-line option as well.

To supply the counters via command-line options, use:

rocprofv3 --pmc SQ_WAVES GRBM_COUNT GRBM_GUI_ACTIVE -- <app_relative_path>

Note

Please note that more than 1 counters should be separated by a space or a comma.
Job will fail if entire set of counters cannot be collected in single pass

Kernel profiling output#

To supply the input file for kernel profiling, use:

rocprofv3 -i input.txt -- <app_relative_path>

Running the above command generates a ./pmc_n/counter_collection.csv file prefixed with the process ID. For each pmc row, a directory pmc_n containing a counter_collection.csv file is generated, where n = 1 for the first row and so on.

In case of JSON or YAML input file, for each job, a directory pass_n containing a counter_collection.csv file is generated where n = 1…N jobs.

Each row of the CSV file is an instance of kernel execution. Here is a truncated version of the output file from pmc_1:

$ cat pmc_1/218_counter_collection.csv

Here are the contents of counter_collection.csv file:

Table 9 Counter collection#
Correlation_Id	Dispatch_Id	Agent_Id	Queue_Id	Process_Id	Thread_Id	Grid_Size	Kernel_Id	Kernel_Name	Workgroup_Size	LDS_Block_Size	Scratch_Size	VGPR_Count	SGPR_Count	Counter_Name	Counter_Value	Start_Timestamp	End_Timestamp
1	1	1	1	19396	19396	1048576	16	void addition_kernel<float>(float, float const, float const*, int, int)	64	0	0	8	16	SQ_WAVES	16384	2228955885095594	2228955885119754
2	2	1	1	19396	19396	1048576	19	subtract_kernel(float, float const, float const*, int, int)	64	0	0	8	16	SQ_WAVES	16384	2228955885095594	2228955885119754
5	5	1	2	19396	19396	1048576	16	void addition_kernel<float>(float, float const, float const*, int, int)	64	0	0	8	16	SQ_WAVES	16384	2228955885095594	2228955885119754
9	9	1	3	19396	19396	1048576	16	void addition_kernel<float>(float, float const, float const*, int, int)	64	0	0	8	16	SQ_WAVES	16384	2228955885095594	2228955885119754
13	13	1	4	19396	19396	1048576	16	void addition_kernel<float>(float, float const, float const*, int, int)	64	0	0	8	16	SQ_WAVES	16384	2228955885095594	2228955885119754
3	3	1	1	19396	19396	1048576	17	multiply_kernel(float, float const, float const*, int, int)	64	0	0	8	16	SQ_WAVES	16384	2228955885095594	2228955885119754
6	6	1	2	19396	19396	1048576	19	subtract_kernel(float, float const, float const*, int, int)	64	0	0	8	16	SQ_WAVES	16384	2228955885095594	2228955885119754

For the description of the fields in the output file, see Output file fields.

Kernel filtering#

rocprofv3 supports kernel filtering in case of profiling. A kernel filter is a set of a regex string (to include the kernels matching this filter), a regex string (to exclude the kernels matching this filter), and an iteration range (set of iterations of the included kernels). If the iteration range is not provided then all iterations of the included kernels are profiled.

$ cat input.yml
jobs:
    - pmc: [SQ_WAVES]
    kernel_include_regex: "divide"
    kernel_exclude_regex: ""
    kernel_iteration_range: "[1, 2, [5-8]]"

Agent info#

Note

All tracing and counter collection options generate an additional agent_info.csv file prefixed with the process ID.

The agent_info.csv file contains information about the CPU or GPU the kernel runs on.

$ cat 238_agent_info.csv

"Node_Id","Logical_Node_Id","Agent_Type","Cpu_Cores_Count","Simd_Count","Cpu_Core_Id_Base","Simd_Id_Base","Max_Waves_Per_Simd","Lds_Size_In_Kb","Gds_Size_In_Kb","Num_Gws","Wave_Front_Size","Num_Xcc","Cu_Count","Array_Count","Num_Shader_Banks","Simd_Arrays_Per_Engine","Cu_Per_Simd_Array","Simd_Per_Cu","Max_Slots_Scratch_Cu","Gfx_Target_Version","Vendor_Id","Device_Id","Location_Id","Domain","Drm_Render_Minor","Num_Sdma_Engines","Num_Sdma_Xgmi_Engines","Num_Sdma_Queues_Per_Engine","Num_Cp_Queues","Max_Engine_Clk_Ccompute","Max_Engine_Clk_Fcompute","Sdma_Fw_Version","Fw_Version","Capability","Cu_Per_Engine","Max_Waves_Per_Cu","Family_Id","Workgroup_Max_Size","Grid_Max_Size","Local_Mem_Size","Hive_Id","Gpu_Id","Workgroup_Max_Dim_X","Workgroup_Max_Dim_Y","Workgroup_Max_Dim_Z","Grid_Max_Dim_X","Grid_Max_Dim_Y","Grid_Max_Dim_Z","Name","Vendor_Name","Product_Name","Model_Name"
0,0,"CPU",24,0,0,0,0,0,0,0,0,1,24,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,3800,0,0,0,0,0,0,23,0,0,0,0,0,0,0,0,0,0,0,"AMD Ryzen 9 3900X 12-Core Processor","CPU","AMD Ryzen 9 3900X 12-Core Processor",""
1,1,"GPU",0,256,0,2147487744,10,64,0,64,64,1,64,4,4,1,16,4,32,90000,4098,26751,12032,0,128,2,0,2,24,3800,1630,432,440,138420864,16,40,141,1024,4294967295,0,0,64700,1024,1024,1024,4294967295,4294967295,4294967295,"gfx900","AMD","Radeon RX Vega","vega10"

Kernel filtering#

Kernel filtering allows you to filter the kernel profiling output based on the kernel name by specifying regex strings in the input file. To include kernel names matching the regex string in the kernel profiling output, use kernel_include_regex. To exclude the kernel names matching the regex string from the kernel profiling output, use kernel_exclude_regex. You can also specify an iteration range for set of iterations of the included kernels. If the iteration range is not specified, then all iterations of the included kernels are profiled.

Here is an input file with kernel filters:

$ cat input.yml
jobs:
    - pmc: [SQ_WAVES]
    kernel_include_regex: "divide"
    kernel_exclude_regex: ""

To collect counters for the kernels matching the filters specified in the preceding input file, run:

rocprofv3 -i input.yml -- <app_relative_path>

$ cat pass_1/312_counter_collection.csv
"Correlation_Id","Dispatch_Id","Agent_Id","Queue_Id","Process_Id","Thread_Id","Grid_Size","Kernel_Name","Workgroup_Size","LDS_Block_Size","Scratch_Size","VGPR_Count","SGPR_Count","Counter_Name","Counter_Value","Start_Timestamp","End_Timestamp"
4,4,1,1,36499,36499,1048576,"divide_kernel(float*, float const*, float const*, int, int)",64,0,0,12,16,"SQ_WAVES",16384,2228955885095594,2228955885119754
8,8,1,2,36499,36499,1048576,"divide_kernel(float*, float const*, float const*, int, int)",64,0,0,12,16,"SQ_WAVES",16384,2228955885095594,2228955885119754
12,12,1,3,36499,36499,1048576,"divide_kernel(float*, float const*, float const*, int, int)",64,0,0,12,16,"SQ_WAVES",16384,2228955892986914,2228955893006114
16,16,1,4,36499,36499,1048576,"divide_kernel(float*, float const*, float const*, int, int)",64,0,0,12,16,"SQ_WAVES",16384,2228955892986914,2228955893006114

Output file fields#

The following table lists the various fields or the columns in the output CSV files generated for application tracing and kernel profiling:

Table 10 output file fields#
Field	Description
Agent_Id	GPU identifier to which the kernel was submitted.
Correlation_Id	Unique identifier for correlation between HIP and HSA async calls during activity tracing.
Start_Timestamp	Begin time in nanoseconds (ns) when the kernel begins execution.
End_Timestamp	End time in ns when the kernel finishes execution.
Queue_Id	ROCm queue unique identifier to which the kernel was submitted.
Private_Segment_Size	The amount of memory required in bytes for the combined private, spill, and arg segments for a work item.
Group_Segment_Size	The group segment memory required by a workgroup in bytes. This does not include any dynamically allocated group segment memory that may be added when the kernel is dispatched.
Workgroup_Size	Size of the workgroup as declared by the compute shader.
Workgroup_Size_n	Size of the workgroup in the nth dimension as declared by the compute shader, where n = X, Y, or Z.
Grid_Size	Number of thread blocks required to launch the kernel.
Grid_Size_n	Number of thread blocks in the nth dimension required to launch the kernel, where n = X, Y, or Z.
LDS_Block_Size	Thread block size for the kernel’s Local Data Share (LDS) memory.
Scratch_Size	Kernel’s scratch memory size.
SGPR_Count	Kernel’s Scalar General Purpose Register (SGPR) count.
VGPR_Count	Kernel’s Vector General Purpose Register (VGPR) count.

Output formats#

rocprofv3 supports the following output formats:

CSV (Default)
JSON (Custom format for programmatic analysis only)
PFTrace (Perfetto trace for visualization with Perfetto)
OTF2 (Open Trace Format for visualization with compatible third party tools)

You can specify the output format using the --output-format command-line option. Format selection is case-insensitive and multiple output formats are supported. For example: --output-format json enables JSON output exclusively whereas --output-format csv json pftrace otf2 enables all four output formats for the run.

For .pftrace trace visualization, use the PFTrace format and open the trace in ui.perfetto.dev.

For .otf2 trace visualization, open the trace in vampir.eu or any supported visualizer.

Note

For large trace files(> 10GB), its recommended to use otf2 format.

JSON output schema#

rocprofv3 supports a custom JSON output format designed for programmatic analysis and NOT for visualization. The schema is optimized for size while factoring in usability. The Perfetto UI does not accept this JSON output format produced by rocprofv3. Perfetto is dropping support for the JSON Chrome tracing format in favor of the binary Perfetto protobuf format (.pftrace extension), which is supported by rocprofv3. You can generate the JSON output using --output-format json command-line option.

Properties#

`rocprofiler-sdk-tool` (array): rocprofv3 data per process (each element represents a process).
- Items (object): Data for rocprofv3.
  
  `metadata` (object, required): Metadata related to the profiler session.
  
  `pid` (integer, required): Process ID.
  
  `init_time` (integer, required): Initialization time in nanoseconds.
  
  `fini_time` (integer, required): Finalization time in nanoseconds.
  
  `agents` (array, required): List of agents.
  
  Items (object): Data for an agent.
  
  `size` (integer, required): Size of the agent data.
  
  `id` (object, required): Identifier for the agent.
  
  `handle` (integer, required): Handle for the agent.
  
  `type` (integer, required): Type of the agent.
  
  `cpu_cores_count` (integer): Number of CPU cores.
  
  `simd_count` (integer): Number of SIMD units.
  
  `mem_banks_count` (integer): Number of memory banks.
  
  `caches_count` (integer): Number of caches.
  
  `io_links_count` (integer): Number of I/O links.
  
  `cpu_core_id_base` (integer): Base ID for CPU cores.
  
  `simd_id_base` (integer): Base ID for SIMD units.
  
  `max_waves_per_simd` (integer): Maximum waves per SIMD.
  
  `lds_size_in_kb` (integer): Size of LDS in KB.
  
  `gds_size_in_kb` (integer): Size of GDS in KB.
  
  `num_gws` (integer): Number of GWS (global work size).
  
  `wave_front_size` (integer): Size of the wave front.
  
  `num_xcc` (integer): Number of XCC (execution compute units).
  
  `cu_count` (integer): Number of compute units (CUs).
  
  `array_count` (integer): Number of arrays.
  
  `num_shader_banks` (integer): Number of shader banks.
  
  `simd_arrays_per_engine` (integer): SIMD arrays per engine.
  
  `cu_per_simd_array` (integer): CUs per SIMD array.
  
  `simd_per_cu` (integer): SIMDs per CU.
  
  `max_slots_scratch_cu` (integer): Maximum slots for scratch CU.
  
  `gfx_target_version` (integer): GFX target version.
  
  `vendor_id` (integer): Vendor ID.
  
  `device_id` (integer): Device ID.
  
  `location_id` (integer): Location ID.
  
  `domain` (integer): Domain identifier.
  
  `drm_render_minor` (integer): DRM render minor version.
  
  `num_sdma_engines` (integer): Number of SDMA engines.
  
  `num_sdma_xgmi_engines` (integer): Number of SDMA XGMI engines.
  
  `num_sdma_queues_per_engine` (integer): Number of SDMA queues per engine.
  
  `num_cp_queues` (integer): Number of CP queues.
  
  `max_engine_clk_ccompute` (integer): Maximum engine clock for compute.
  
  `max_engine_clk_fcompute` (integer): Maximum engine clock for F compute.
  
  `sdma_fw_version` (object): SDMA firmware version.
  
  `uCodeSDMA` (integer, required): SDMA microcode version.
  
  `uCodeRes` (integer, required): Reserved microcode version.
  
  `fw_version` (object): Firmware version.
  
  `uCode` (integer, required): Microcode version.
  
  `Major` (integer, required): Major version.
  
  `Minor` (integer, required): Minor version.
  
  `Stepping` (integer, required): Stepping version.
  
  `capability` (object, required): Agent capability flags.
  
  `HotPluggable` (integer, required): Hot pluggable capability.
  
  `HSAMMUPresent` (integer, required): HSAMMU present capability.
  
  `SharedWithGraphics` (integer, required): Shared with graphics capability.
  
  `QueueSizePowerOfTwo` (integer, required): Queue size is power of two.
  
  `QueueSize32bit` (integer, required): Queue size is 32-bit.
  
  `QueueIdleEvent` (integer, required): Queue idle event.
  
  `VALimit` (integer, required): VA limit.
  
  `WatchPointsSupported` (integer, required): Watch points supported.
  
  `WatchPointsTotalBits` (integer, required): Total bits for watch points.
  
  `DoorbellType` (integer, required): Doorbell type.
  
  `AQLQueueDoubleMap` (integer, required): AQL queue double map.
  
  `DebugTrapSupported` (integer, required): Debug trap supported.
  
  `WaveLaunchTrapOverrideSupported` (integer, required): Wave launch trap override supported.
  
  `WaveLaunchModeSupported` (integer, required): Wave launch mode supported.
  
  `PreciseMemoryOperationsSupported` (integer, required): Precise memory operations supported.
  
  `DEPRECATED_SRAM_EDCSupport` (integer, required): Deprecated SRAM EDC support.
  
  `Mem_EDCSupport` (integer, required): Memory EDC support.
  
  `RASEventNotify` (integer, required): RAS event notify.
  
  `ASICRevision` (integer, required): ASIC revision.
  
  `SRAM_EDCSupport` (integer, required): SRAM EDC support.
  
  `SVMAPISupported` (integer, required): SVM API supported.
  
  `CoherentHostAccess` (integer, required): Coherent host access.
  
  `DebugSupportedFirmware` (integer, required): Debug supported firmware.
  
  `Reserved` (integer, required): Reserved field.
  
  `counters` (array, required): Array of counter objects.
  
  Items (object)
  
  `agent_id` (object, required): Agent ID information.
  
  `handle` (integer, required): Handle of the agent.
  
  `id` (object, required): Counter ID information.
  
  `handle` (integer, required): Handle of the counter.
  
  `is_constant` (integer, required): Indicator if the counter value is constant.
  
  `is_derived` (integer, required): Indicator if the counter value is derived.
  
  `name` (string, required): Name of the counter.
  
  `description` (string, required): Description of the counter.
  
  `block` (string, required): Block information of the counter.
  
  `expression` (string, required): Expression of the counter.
  
  `dimension_ids` (array, required): Array of dimension IDs.
  
  Items (integer): Dimension ID.
  
  `strings` (object, required): String records.
  
  `callback_records` (array): Callback records.
  
  Items (object)
  
  `kind` (string, required): Kind of the record.
  
  `operations` (array, required): Array of operations.
  
  Items (string): Operation.
  
  `buffer_records` (array): Buffer records.
  
  Items (object)
  
  `kind` (string, required): Kind of the record.
  
  `operations` (array, required): Array of operations.
  
  Items (string): Operation.
  
  `marker_api` (array): Marker API records.
  
  Items (object)
  
  `key` (integer, required): Key of the record.
  
  `value` (string, required): Value of the record.
  
  `counters` (object): Counter records.
  
  `dimension_ids` (array, required): Array of dimension IDs.
  
  Items (object)
  
  `id` (integer, required): Dimension ID.
  
  `instance_size` (integer, required): Size of the instance.
  
  `name` (string, required): Name of the dimension.
  
  `code_objects` (array, required): Code object records.
  
  Items (object)
  
  `size` (integer, required): Size of the code object.
  
  `code_object_id` (integer, required): ID of the code object.
  
  `rocp_agent` (object, required): ROCP agent information.
  
  `handle` (integer, required): Handle of the ROCP agent.
  
  `hsa_agent` (object, required): HSA agent information.
  
  `handle` (integer, required): Handle of the HSA agent.
  
  `uri` (string, required): URI of the code object.
  
  `load_base` (integer, required): Base address for loading.
  
  `load_size` (integer, required): Size for loading.
  
  `load_delta` (integer, required): Delta for loading.
  
  `storage_type` (integer, required): Type of storage.
  
  `memory_base` (integer, required): Base address for memory.
  
  `memory_size` (integer, required): Size of memory.
  
  `kernel_symbols` (array, required): Kernel symbol records.
  
  Items (object)
  
  `size` (integer, required): Size of the kernel symbol.
  
  `kernel_id` (integer, required): ID of the kernel.
  
  `code_object_id` (integer, required): ID of the code object.
  
  `kernel_name` (string, required): Name of the kernel.
  
  `kernel_object` (integer, required): Object of the kernel.
  
  `kernarg_segment_size` (integer, required): Size of the kernarg segment.
  
  `kernarg_segment_alignment` (integer, required): Alignment of the kernarg segment.
  
  `group_segment_size` (integer, required): Size of the group segment.
  
  `private_segment_size` (integer, required): Size of the private segment.
  
  `formatted_kernel_name` (string, required): Formatted name of the kernel.
  
  `demangled_kernel_name` (string, required): Demangled name of the kernel.
  
  `truncated_kernel_name` (string, required): Truncated name of the kernel.
  
  `callback_records` (object, required): Callback record details.
  
  `counter_collection` (array): Counter collection records.
  
  Items (object)
  
  `dispatch_data` (object, required): Dispatch data details.
  
  `size` (integer, required): Size of the dispatch data.
  
  `correlation_id` (object, required): Correlation ID information.
  
  `internal` (integer, required): Internal correlation ID.
  
  `external` (integer, required): External correlation ID.
  
  `dispatch_info` (object, required): Dispatch information details.
  
  `size` (integer, required): Size of the dispatch information.
  
  `agent_id` (object, required): Agent ID information.
  
  `handle` (integer, required): Handle of the agent.
  
  `queue_id` (object, required): Queue ID information.
  
  `handle` (integer, required): Handle of the queue.
  
  `kernel_id` (integer, required): ID of the kernel.
  
  `dispatch_id` (integer, required): ID of the dispatch.
  
  `private_segment_size` (integer, required): Size of the private segment.
  
  `group_segment_size` (integer, required): Size of the group segment.
  
  `workgroup_size` (object, required): Workgroup size information.
  
  `x` (integer, required): X dimension.
  
  `y` (integer, required): Y dimension.
  
  `z` (integer, required): Z dimension.
  
  `grid_size` (object, required): Grid size information.
  
  `x` (integer, required): X dimension.
  
  `y` (integer, required): Y dimension.
  
  `z` (integer, required): Z dimension.
  
  `records` (array, required): Records.
  
  Items (object)
  
  `counter_id` (object, required): Counter ID information.
  
  `handle` (integer, required): Handle of the counter.
  
  `value` (number, required): Value of the counter.
  
  `thread_id` (integer, required): Thread ID.
  
  `arch_vgpr_count` (integer, required): Count of VGPRs.
  
  `sgpr_count` (integer, required): Count of SGPRs.
  
  `lds_block_size_v` (integer, required): Size of LDS block.
  
  `buffer_records` (object, required): Buffer record details.
  
  `kernel_dispatch` (array): Kernel dispatch records.
  
  Items (object)
  
  `size` (integer, required): Size of the dispatch.
  
  `kind` (integer, required): Kind of the dispatch.
  
  `operation` (integer, required): Operation of the dispatch.
  
  `thread_id` (integer, required): Thread ID.
  
  `correlation_id` (object, required): Correlation ID information.
  
  `internal` (integer, required): Internal correlation ID.
  
  `external` (integer, required): External correlation ID.
  
  `start_timestamp` (integer, required): Start timestamp.
  
  `end_timestamp` (integer, required): End timestamp.
  
  `dispatch_info` (object, required): Dispatch information details.
  
  `size` (integer, required): Size of the dispatch information.
  
  `agent_id` (object, required): Agent ID information.
  
  `handle` (integer, required): Handle of the agent.
  
  `queue_id` (object, required): Queue ID information.
  
  `handle` (integer, required): Handle of the queue.
  
  `kernel_id` (integer, required): ID of the kernel.
  
  `dispatch_id` (integer, required): ID of the dispatch.
  
  `private_segment_size` (integer, required): Size of the private segment.
  
  `group_segment_size` (integer, required): Size of the group segment.
  
  `workgroup_size` (object, required): Workgroup size information.
  
  `x` (integer, required): X dimension.
  
  `y` (integer, required): Y dimension.
  
  `z` (integer, required): Z dimension.
  
  `grid_size` (object, required): Grid size information.
  
  `x` (integer, required): X dimension.
  
  `y` (integer, required): Y dimension.
  
  `z` (integer, required): Z dimension.
  
  `hip_api` (array): HIP API records.
  
  Items (object)
  
  `size` (integer, required): Size of the HIP API record.
  
  `kind` (integer, required): Kind of the HIP API.
  
  `operation` (integer, required): Operation of the HIP API.
  
  `correlation_id` (object, required): Correlation ID information.
  
  `internal` (integer, required): Internal correlation ID.
  
  `external` (integer, required): External correlation ID.
  
  `start_timestamp` (integer, required): Start timestamp.
  
  `end_timestamp` (integer, required): End timestamp.
  
  `thread_id` (integer, required): Thread ID.
  
  `hsa_api` (array): HSA API records.
  
  Items (object)
  
  `size` (integer, required): Size of the HSA API record.
  
  `kind` (integer, required): Kind of the HSA API.
  
  `operation` (integer, required): Operation of the HSA API.
  
  `correlation_id` (object, required): Correlation ID information.
  
  `internal` (integer, required): Internal correlation ID.
  
  `external` (integer, required): External correlation ID.
  
  `start_timestamp` (integer, required): Start timestamp.
  
  `end_timestamp` (integer, required): End timestamp.
  
  `thread_id` (integer, required): Thread ID.
  
  `marker_api` (array): Marker (ROCTx) API records.
  
  Items (object)
  
  `size` (integer, required): Size of the Marker API record.
  
  `kind` (integer, required): Kind of the Marker API.
  
  `operation` (integer, required): Operation of the Marker API.
  
  `correlation_id` (object, required): Correlation ID information.
  
  `internal` (integer, required): Internal correlation ID.
  
  `external` (integer, required): External correlation ID.
  
  `start_timestamp` (integer, required): Start timestamp.
  
  `end_timestamp` (integer, required): End timestamp.
  
  `thread_id` (integer, required): Thread ID.
  
  `memory_copy` (array): Async memory copy records.
  
  Items (object)
  
  `size` (integer, required): Size of the Marker API record.
  
  `kind` (integer, required): Kind of the Marker API.
  
  `operation` (integer, required): Operation of the Marker API.
  
  `correlation_id` (object, required): Correlation ID information.
  
  `internal` (integer, required): Internal correlation ID.
  
  `external` (integer, required): External correlation ID.
  
  `start_timestamp` (integer, required): Start timestamp.
  
  `end_timestamp` (integer, required): End timestamp.
  
  `thread_id` (integer, required): Thread ID.
  
  `dst_agent_id` (object, required): Destination Agent ID.
  
  `handle` (integer, required): Handle of the agent.
  
  `src_agent_id` (object, required): Source Agent ID.
  
  `handle` (integer, required): Handle of the agent.
  
  `bytes` (integer, required): Bytes copied.

Using rocprofv3

Contents

Using rocprofv3#

Options#

Application tracing#

HIP trace#

HSA trace#

Marker trace#

Kernel trace#

Memory copy trace#

Runtime trace#

System trace#

Scratch memory trace#

Stats#

Kernel profiling#

Input file#

Properties#

Command-line#

Kernel profiling output#

Kernel filtering#

Agent info#

Kernel filtering#

Output file fields#

Output formats#

JSON output schema#

Properties#