Commit Graph

13 Commits

Author SHA1 Message Date
Jonathan R. Madsen d93d4d9463 rocprofv3: support specifying HW counters via command line (#1130)
* rocprofv3: support specifying PMC counters via command line

- E.g. `rocprofv3 --pmc SQ_WAVES -- <app>`

* Update CHANGELOG

* updated rocprofv3 help and documentation

---------

Co-authored-by: Gopesh Bhardwaj <gopesh.bhardwaj@amd.com>
2024-10-25 02:49:30 -05:00
Jonathan R. Madsen a92fa8f071 Miscellaneous Testing Fixes + CMake find_package include guard (#1106)
* Improve ROCPROFILER_DEFAULT_FAIL_REGEX

* Support find_package called twice

* Skip iteration of ROCPROFILER_AGENT_TYPE_CPU

* Relax tests/rocprofv3/summary

* Tweak to rocprofiler-sdk-tool/tool.cpp

* Move rocprofv3-trigger-list-metrics to libexec

* Remove PC_SAMPLING_TESTS_REGEX from CI workflow

* Update packaging

- core component depends on roctx component

* Increase verbosity for failing tests

* Fix RPATH of rocprofv3-trigger-list-metrics

---------

Co-authored-by: Gopesh Bhardwaj <gopesh.bhardwaj@amd.com>
2024-10-22 16:24:16 +05:30
Gopesh Bhardwaj 320427b5f5 rocprofv3: docs and help menu updates (#1129)
* doc updates

* Correcting ROCtx information

* Making ROCTx string consistent

* missing occurence
2024-10-17 13:28:53 +05:30
Gopesh Bhardwaj 9b2ece76c3 Fixing installed pacakge tests in CI (#1119)
* Fixing installed pacakge tests in CI

* Formatted rocprofv3.py with black formatter
2024-10-07 20:55:35 +05:30
Gopesh Bhardwaj f2f267bcc0 updating rocprofv3 help options (#1113)
* updating rocprofv3 help options

* updating CHANGELOG
2024-10-05 12:40:27 +05:30
Jonathan R. Madsen d5bcb63263 rocprofv3 Kokkos-Tools Support (#1058) 2024-09-12 00:46:07 -05:00
Mythreya 2a146259c7 Add support for RCCL tracing (#1047)
* [Draft]: Add support for RCCL tracing

Address comments

* [Draft]: Add support for RCCL tracing

Address PR comments, changes from RCCL upstream

* Add RCCL library table registration

Working on adding support to rocprofiler-register

* Support compilation w/o <rccl/amd_detail/api_trace.h>

- dummy api_trace.h header
- return ROCPROFILER_STATUS_ERROR_NOT_IMPLEMENTED when RCCL does not have api_trace.h header

* RCCL API tracing tool support

- add to rocprofv3
- add to json-tool

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
2024-09-12 00:42:58 -05:00
Jonathan R. Madsen 72cbcedc9e LD_PRELOAD librocprofiler-sdk-roctx.so when marker-trace enabled (#1057)
* LD_PRELOAD librocprofiler-sdk-roctx.so when marker-trace enabled

- this enables apps to link against old ROCTx (libroctx64.so) but get marker tracing in rocprofv3

* Update CHANGELOG

* Validation test for app linked to old (roctracer) ROCTx library

* Tweak scope of tool_counter_info

- causing "signal-unsafe call inside of a signal" error for ThreadSanitizer on mi200

* Fix handling of missing transpose-roctracer-roctx

* Disable rocprofv3 aborted-app test (ThreadSanitizer)

- ThreadSanitizer + mi200/mi300 + aborted-app results in a signal-unsafe call inside a signal that cannot be specifically suppressed as usual via rocprofv3_error_signal_handler for some unknown reason

* Add UndefinedBehaviorSanitizer job
2024-09-11 15:27:35 -05:00
Jonathan R. Madsen 395f01b689 rocprofv3: summary reports + more JSON metadata (#1029)
* Move include/rocprofiler-sdk/cxx/details/delimit.hpp to tokenize.hpp

* Update docs/how-to/using-rocprofv3.rst

- fix code block indents
- reorder rocprofv3 options, limit them to important options
- add docs for `--runtime-trace`

* Update rocprofv3.py

- parser argument groups
- new `--runtime-trace` option
- new `--summary` option
- new `--summary-per-domain` option
- new `--summary-groups` option
- new `--summary-output-file` option
- new `--summary-units` option

* Update lib/rocprofiler-sdk/hsa/async_copy.cpp

- fix async copy operation names: add "MEMORY_COPY_" prefix

* lib/rocprofiler-sdk-tool: update statistics.{hpp,cpp}

- statistics<>::get_percent function
- stats_entry_t struct
- stats_formatter struct
- percentage struct
- std::to_string(::rocprofiler::tool::percentage)

* lib/rocprofiler-sdk-tool: update domain_type.{hpp,cpp}

- reorder domain_type enum values

* lib/rocprofiler-sdk-tool: update generateCSV.{hpp,cpp}

- separate writing CSV from accumulating statistics
- a lot of functionality was moved to statistics.{hpp,cpp}

* lib/rocprofiler-sdk-tool: update output_file.{hpp,cpp}

- output_stream_t struct
- get_output_stream(...) returns output_stream_t instance

* lib/rocprofiler-sdk-tool: update generateJSON.cpp

- update get_output_stream usage to output_stream_t

* lib/rocprofiler-sdk-tool: update generateOTF2.cpp

- header include order tweak

* lib/rocprofiler-sdk-tool: update buffered_output.hpp

- stats_data_t was renamed to stats_entry_t

* lib/rocprofiler-sdk-tool: update generatePerfetto.cpp

- header include tweak

* lib/rocprofiler-sdk-tool: update tmp_file_buffer.hpp

- emit warning message if write_ring_buffer fails after offloading instead of aborting
- prefer placement new instead of assignment in write_ring_buffer

* lib/rocprofiler-sdk-tool: add generateStats.{hpp,cpp}

- functions for accumulating statistics

* Update tests/rocprofv3/tracing-hip-in-libraries/CMakeLists.txt

- accommodate tweak to CSV output file name for HIP and HSA traces

* lib/rocprofiler-sdk-tool: update config.{hpp,cpp}

- new config variables
  - stats_summary
  - stats_summary_per_domain
  - summary_output
  - stats_summary_unit_value
  - stats_summary_unit
  - stats_summary_file
  - stats_summary_groups
- support output keys for hostname: %hostname% / %h

* lib/rocprofiler-sdk-tool: update tool.cpp

- support summary output

* Documentation fixes

* Test for summary output

* Update tests/bin/transpose to use more ROCTx

- also support building with the roctracer ROCTx

* Remove roctxMark from OTF2 + fix kernel-rename tests

- following more ROCTx calls in transpose, kernel-rename validation had to be updated

* JSON metadata + JSON summary

- add serialization support for config
- add serialization support for statistics
- additions to json spec
  - rocprofiler-sdk-tool/metadata/config
  - rocprofiler-sdk-tool/metadata/command
  - rocprofiler-sdk-tool/summary
- config output_keys support for NVIDIA %q{<ENV-VAR>} syntax
- config output_keys support keys within keys

* rocprofv3 --summary-groups warning if no domain matches

- emit warning if a regex in for summary groups did not match any domain names

* Compile fix for lib/rocprofiler-sdk-tool/tool.cpp

- get_config().scratch_memory_trace
- pass contributions to write_json

* Update rocprofv3.py to preload rocprofiler-sdk-roctx

- appended to LD_PRELOAD when args.marker_trace is enabled

* Fix ReST link errors about subtitle underline being too short

* Patch tokenization of config::stats_summary_groups

- guard against array values of empty strings

* Tweak rocprofv3 summary test

- input-summary.yaml (used by rocprofv3-test-summary-inp-yaml-execute) only provides one summary group regex

* Disable LD_PRELOAD of librocprofiler-sdk-roctx.so

- this causes problems in the sanitizers, will be addressed in another PR
2024-09-09 11:20:55 -05:00
Jonathan R. Madsen 16d535ef48 rocprofv3 OTF2 Output Support (#995)
* CMake support for OTF2 library

* Preliminary OTF2 generation implementation

* Completed OTF2 Support

- HSA API
- HIP API
- Marker API
- Async Memory Copies
- Kernel Dispatch

* Update lib/rocprofiler-sdk-tool/generateOTF2.cpp

- fix location type for dispatches

* Testing for OTF2 output

* Add OTF2 to requirements.txt

* Update lib/rocprofiler-sdk-tool/generateOTF2.cpp

- fix getting kernel name

* OTF2 testing with rocprofv3/tracing-hip-in-libraries

* Format external/otf2/CMakeLists.txt

* Update external/otf2/CMakeLists.txt

- guard CMP0135 for cmake < 3.24

* Update lib/rocprofiler-sdk-tool/generateOTF2.cpp

- fix duplicate string ref issue

* Update lib/rocprofiler-sdk-tool/generateOTF2.cpp

- fix header includes

* Update CI workflow

- sudo install pypi requirements for core-rpm for $HOME/.local installs

* Update pytest_utils/otf2_reader.py

- modifications for reading trace

* Update pytest_utils/otf2_reader.py

- misc cleanup

* Update CI workflow

- fix installer artifact naming

* Update pytest_utils/otf2_reader.py

- handle slightly overlapping kernel timestamps for MI300

* OTF2 attributes for category

* Testing with OTF2Reader category attributes

* Fix memory leak in OTF2 generation

- leaking OTF2_AttributeList
2024-07-30 19:57:19 -05:00
Jonathan R. Madsen ebb021c59f rocprofv3 kernel renaming support + misc rocprofv3 updates (#992)
* Increase rocprofv3 tool buffer size

- 32 pages instead of 1 page

* Improve rocprofv3 perfetto track labels

* Preliminary kernel renaming support + misc rocprofv3 updates

- add rocprofv3 option --kernel-rename
- add rocprofv3 options for perfetto settings (buffer size, etc.)
- add CSV columns for kernel trace
  - Thread_Id
  - Dispatch_Id
- add CSV column for counter_collection
 - Kernel_Id
2024-07-29 14:33:50 -05:00
SrirakshaNag ace34abd11 kernel iteration filtering for counter collection (#911)
* kernel filtering for counter collection

* fixing trace tests

* removing print statements

* fix CI fail

* handling preload and updating docs

* minor fix

* misc fix

* misc fix

* Typo fix

* Update rocprofv3 + input schema

- "application_passes" -> "jobs"
- removed nesting in YAML/JSON inputs
- improved customAction (now booleanArgAction)
  - supports --<name> (defaults to true)
  - supports --<name>=<truth-value>
  - supports --<name> <truth-value>
- added --kernel-iteration-range to command-line
- automatically support new command-line options in YAML/JSON input
- standardized PMC return from text input to match PMC from YAML/JSON input
- added support for --log-level env
- updated various input*.(yml|json) to modified schema

* Update config.cpp

- added recommended code to get_kernel_filter_range

* Fixing iteration

* misc fix

* support only [-] for iteration

* bug fix

* Fix using-rocprofv3.rst

* Update config.cpp

- patch get_kernel_filter_range

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
2024-07-26 21:46:53 -05:00
Gopesh Bhardwaj dc054eea76 Adding python script for rocprofv3 (#849)
* Adding python script for rocprofv3

* script update

* updating script

* Fixing script for counter collection tests

* Fixing Sanitizer issues

* Adding YAML input file support

* Fixing counter validation tests

* script modifications

* Adding missing LD_PRELOAD

* doc updates

* Adding test for yaml input

* updated yaml extension support in doc

* backward compatibility

* updating scripts

* Fixing git history rocprofv3- part1

* Fixing git history rocprofv3- part2

* Fixing rocprofv3 history final

* Rebasing PR 860 for json support

* Review comments: Parser updates

* Removed color encoding and rebasing again

* Addressing review comments

* removing globals

* Update rocprofv3.py

- update tests to conform to new argparse requirements
- added support for JSON

* Slight tweak to update_env for ROCP_OUTPUT_FORMAT

* Update rocprofv3.py

- Handle ROCPROF_PRELOAD
- Add --preload option
- Add --kernel-names option

* Update rocprofv3.py

- Fix update_env handling of prepend/append
- Tweak --preload argument

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
2024-05-29 05:05:10 -05:00