Commit Graph

64650 Commits

Author SHA1 Message Date
anujshuk-amd 35b8748c20 Fix ROCPROFSYS_AMD_SMI_METRICS parsing (#178)
Fixes a bug where all the `ROCPROFSYSE_AMD_SMI_METRICS` values were being recorded by default.
Fixes bug with the 'all' and 'none' values giving an exception when specified for `ROCPROFSYSE_AMD_SMI_METRICS`.

---------

Co-authored-by: David Galiffi <David.Galiffi@amd.com>

[ROCm/rocprofiler-systems commit: 8d48048bd3]
2025-04-28 09:22:20 -04:00
anujshuk-amd 8d48048bd3 Fix ROCPROFSYS_AMD_SMI_METRICS parsing (#178)
Fixes a bug where all the `ROCPROFSYSE_AMD_SMI_METRICS` values were being recorded by default.
Fixes bug with the 'all' and 'none' values giving an exception when specified for `ROCPROFSYSE_AMD_SMI_METRICS`.

---------

Co-authored-by: David Galiffi <David.Galiffi@amd.com>
2025-04-28 09:22:20 -04:00
Godavarthy Surya, Anusha ff69bcc903 SWDEV-469422 - Avoid using of hipStream_t in internal methods (#69)
Change-Id: Ifd5362f371c846a88241927383cb95cf046548ef

[ROCm/clr commit: fb92683d86]
2025-04-28 15:09:11 +05:30
Godavarthy Surya, Anusha fb92683d86 SWDEV-469422 - Avoid using of hipStream_t in internal methods (#69)
Change-Id: Ifd5362f371c846a88241927383cb95cf046548ef
2025-04-28 15:09:11 +05:30
Godavarthy Surya, Anusha 0eb2e5e8f2 SWDEV-469422 - hipGraphNodeDOTAttribute change std::string members to const char* (#70)
Compiler creates global variables for every unique string

Change-Id: I4cf8dd3e763d16740096e345da67a7ef72f61515

[ROCm/clr commit: bbcb1f9c70]
2025-04-28 14:57:36 +05:30
Godavarthy Surya, Anusha bbcb1f9c70 SWDEV-469422 - hipGraphNodeDOTAttribute change std::string members to const char* (#70)
Compiler creates global variables for every unique string

Change-Id: I4cf8dd3e763d16740096e345da67a7ef72f61515
2025-04-28 14:57:36 +05:30
Madsen, Jonathan 394d3dcf45 Fix evaulate_ast.evaulate_hybrid_counters test (#377)
- broken by #348

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>

[ROCm/rocprofiler-sdk commit: 032c06db9a]
2025-04-27 22:42:10 -05:00
Madsen, Jonathan 032c06db9a Fix evaulate_ast.evaulate_hybrid_counters test (#377)
- broken by #348

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
2025-04-27 22:42:10 -05:00
Aleksandar Djordjevic 395bf369fc [CMake] Fix GPU detection function
* Fix cmake GPU detection

* formatting cmake

---------

Co-authored-by: David Galiffi <David.Galiffi@amd.com>

[ROCm/rocprofiler-systems commit: 39f676aa71]
2025-04-28 05:07:00 +02:00
Aleksandar Djordjevic 39f676aa71 [CMake] Fix GPU detection function
* Fix cmake GPU detection

* formatting cmake

---------

Co-authored-by: David Galiffi <David.Galiffi@amd.com>
2025-04-28 05:07:00 +02:00
Madsen, Jonathan 871686ad40 [rocprofv3] Use -P for collection period shorthand option (#356)
* [rocprofv3] Use -P for collection period option

- Reserve -p for profiler attachment

* Update changelog

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>

[ROCm/rocprofiler-sdk commit: d2bde3ce27]
2025-04-27 20:18:26 -05:00
Madsen, Jonathan d2bde3ce27 [rocprofv3] Use -P for collection period shorthand option (#356)
* [rocprofv3] Use -P for collection period option

- Reserve -p for profiler attachment

* Update changelog

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
2025-04-27 20:18:26 -05:00
Madsen, Jonathan 9f7703f918 Build system (libdw), correlation ID, and shebang fixes (#354)
* Fix compilation for output library

- link to targets for ATT (amd-comgr, dw, elf)

* Relax correlation ID retirement log failures

- only fail for correlation ID retirement underflow when building in CI mode

* Fix shebang for several files

- license was inserted before shebang in several places

* Update code coverage exclude folders for samples

* Tweak to agent tests

- test to make sure hsa agent is not the old value instead of testing that it is the new value

* Fix libdw include/link

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>

[ROCm/rocprofiler-sdk commit: 3580478426]
2025-04-27 20:16:18 -05:00
Madsen, Jonathan 3580478426 Build system (libdw), correlation ID, and shebang fixes (#354)
* Fix compilation for output library

- link to targets for ATT (amd-comgr, dw, elf)

* Relax correlation ID retirement log failures

- only fail for correlation ID retirement underflow when building in CI mode

* Fix shebang for several files

- license was inserted before shebang in several places

* Update code coverage exclude folders for samples

* Tweak to agent tests

- test to make sure hsa agent is not the old value instead of testing that it is the new value

* Fix libdw include/link

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
2025-04-27 20:16:18 -05:00
Kandula, Venkateshwar reddy 2b8cbbf711 [SDK] SWDEV-524163 - Add error msg when accumulate is used on counters not from sq block. (#348)
* add error log when accumulate is used on counters not from sq block.

* Address comments.

---------

Co-authored-by: Venkateshwar Reddy Kandula <vkandula@amd.com>

[ROCm/rocprofiler-sdk commit: 06005c7f6b]
2025-04-27 20:03:27 -05:00
Kandula, Venkateshwar reddy 06005c7f6b [SDK] SWDEV-524163 - Add error msg when accumulate is used on counters not from sq block. (#348)
* add error log when accumulate is used on counters not from sq block.

* Address comments.

---------

Co-authored-by: Venkateshwar Reddy Kandula <vkandula@amd.com>
2025-04-27 20:03:27 -05:00
Nagaraj, Sriraksha a83ee16d51 [Misc] fix the agent_id field (#297)
* fix the agent_id field

* Fix shebang

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>

[ROCm/rocprofiler-sdk commit: 35069a9d06]
2025-04-27 19:56:45 -05:00
Nagaraj, Sriraksha 35069a9d06 [Misc] fix the agent_id field (#297)
* fix the agent_id field

* Fix shebang

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
2025-04-27 19:56:45 -05:00
Meserve, Mark d10a9d855a [SDK] Fix for hang in hsa_barrier (#316)
* SWDEV-515895: Fix for hang in hsa_barrier

- Root cause of hang is enqueue_packet using a completed barrier
  - This occurs because set_barrier does not correctly mark the barrier
    as complete.
- Fixed by consolidating duplicate code for how a barrier is:
  - Marked as complete
    - Added common function and trace message
  - Checked as complete
    - Removed internal completion tracking variable in favor of signal
- Added some log messages which were useful in tracking this hang
- Destroying an hsa_barrier now calls the provided completion callback

* formatting

* Fix hsa_barrier test

- Removes assumption that enqueue_packet always generates packets
- Add assertion that a complete barrier does not generate packets

* formatting

* Address review comments

[ROCm/rocprofiler-sdk commit: 61307442f0]
2025-04-27 19:55:32 -05:00
Meserve, Mark 61307442f0 [SDK] Fix for hang in hsa_barrier (#316)
* SWDEV-515895: Fix for hang in hsa_barrier

- Root cause of hang is enqueue_packet using a completed barrier
  - This occurs because set_barrier does not correctly mark the barrier
    as complete.
- Fixed by consolidating duplicate code for how a barrier is:
  - Marked as complete
    - Added common function and trace message
  - Checked as complete
    - Removed internal completion tracking variable in favor of signal
- Added some log messages which were useful in tracking this hang
- Destroying an hsa_barrier now calls the provided completion callback

* formatting

* Fix hsa_barrier test

- Removes assumption that enqueue_packet always generates packets
- Add assertion that a complete barrier does not generate packets

* formatting

* Address review comments
2025-04-27 19:55:32 -05:00
Assiouras, Ioannis 875468bbfb SWDEV-526188 - Fix race condition in StatCO::getStatFunc()
Make sure that a newly created FatBinaryInfo is assigned to modules only after extractFatBinary has been called for the object.

[ROCm/clr commit: 1099e0a131]
2025-04-27 21:14:01 +01:00
Assiouras, Ioannis 1099e0a131 SWDEV-526188 - Fix race condition in StatCO::getStatFunc()
Make sure that a newly created FatBinaryInfo is assigned to modules only after extractFatBinary has been called for the object.
2025-04-27 21:14:01 +01:00
Baraldi, Giovanni f91f0712f7 SWDEV-528686: ATT fix for gfx12 s_wait_idle. Fixes for csv. Default to parse to trace. Fix for ROCR_VISIBLE_DEVICES. (#345)
* Fix for gfx12 s_wait_idle. Added wait field on att.csv

* Format and default to ATT to trace

* Update .mds

* No fatal error for invalid agent

* Tidy fixes

* Rename wait to idle, removed uneeded headers

* Remove unused traceID

* Tidy fix

* Fix csv output

* Formatting

* Fix tests

* Fix tests

* Fix for visible devices

* Review comment: Fix cmake

* Review suggestion

* Remove changelog/readme

* Review comments

* Review comment for CSV

* Formatting

---------

Co-authored-by: Giovanni Baraldi <gbaraldi@amd.com>

[ROCm/rocprofiler-sdk commit: a8f3397069]
2025-04-25 11:49:16 -05:00
Baraldi, Giovanni a8f3397069 SWDEV-528686: ATT fix for gfx12 s_wait_idle. Fixes for csv. Default to parse to trace. Fix for ROCR_VISIBLE_DEVICES. (#345)
* Fix for gfx12 s_wait_idle. Added wait field on att.csv

* Format and default to ATT to trace

* Update .mds

* No fatal error for invalid agent

* Tidy fixes

* Rename wait to idle, removed uneeded headers

* Remove unused traceID

* Tidy fix

* Fix csv output

* Formatting

* Fix tests

* Fix tests

* Fix for visible devices

* Review comment: Fix cmake

* Review suggestion

* Remove changelog/readme

* Review comments

* Review comment for CSV

* Formatting

---------

Co-authored-by: Giovanni Baraldi <gbaraldi@amd.com>
2025-04-25 11:49:16 -05:00
Amber Lin 5e28208cec kfdtest: Skip SVMEvict with xnack=0
Random driver deadlock on svm_range_evict_svm_bo_worker() is obeserved on
NPS2/DPX mode. It's seen with xnack off and happens more often on the
partition with less VRAM because of TMR.

Temporarily skip SVM Evict tests on Family AV when xnack is disabled.

Signed-off-by: Amber Lin <Amber.Lin@amd.com>
2025-04-25 12:45:36 -04:00
Amber Lin 9d98d7479d kfdtest: Skip SVMEvict with xnack=0
Random driver deadlock on svm_range_evict_svm_bo_worker() is obeserved on
NPS2/DPX mode. It's seen with xnack off and happens more often on the
partition with less VRAM because of TMR.

Temporarily skip SVM Evict tests on Family AV when xnack is disabled.

Signed-off-by: Amber Lin <Amber.Lin@amd.com>


[ROCm/ROCR-Runtime commit: 5e28208cec]
2025-04-25 12:45:36 -04:00
Kuricheti, Mythreya cd106cda3c [CI]: test-async-copy-tracing-validate : test_ancestor_ids (#364)
* Add ancestor ID to pftrace

* Simplify hipMemcpyAsync ancestor ID checks

* Review comments

[ROCm/rocprofiler-sdk commit: cac3ac1b98]
2025-04-25 11:19:17 -05:00
Kuricheti, Mythreya cac3ac1b98 [CI]: test-async-copy-tracing-validate : test_ancestor_ids (#364)
* Add ancestor ID to pftrace

* Simplify hipMemcpyAsync ancestor ID checks

* Review comments
2025-04-25 11:19:17 -05:00
Kudchadker, Saleel cd14def193 SWDEV-521647 - Fix tracking of hw_event (#206)
- When a command may possibly have two packets(like device heap
  initializer), and if there is no signal on the main kernel packet the
tracking was broken as it marked HW event of the command as the first
packet signal.
- Make sure if no completion signal is attached to the second packet
  then clear the HW event for the command.

[ROCm/clr commit: 072fb0804e]
2025-04-25 08:46:44 -07:00
Kudchadker, Saleel 072fb0804e SWDEV-521647 - Fix tracking of hw_event (#206)
- When a command may possibly have two packets(like device heap
  initializer), and if there is no signal on the main kernel packet the
tracking was broken as it marked HW event of the command as the first
packet signal.
- Make sure if no completion signal is attached to the second packet
  then clear the HW event for the command.
2025-04-25 08:46:44 -07:00
Kudchadker, Saleel 1b1d6b841e SWDEV-510186 - Improve logging (#220)
- Print all arguments for logs, this is useful for debug

[ROCm/clr commit: ce24936970]
2025-04-25 08:40:31 -07:00
Kudchadker, Saleel ce24936970 SWDEV-510186 - Improve logging (#220)
- Print all arguments for logs, this is useful for debug
2025-04-25 08:40:31 -07:00
Kandula, Venkateshwar reddy 1264ffeb45 [PSDB] [CI] SWDEV-528922: Modify summary MEMORY_ALLOCATION kind tests to account new hip runtime memory manger (#368)
* Modify summary MEMORY_ALLOCATION kind tests to account new hip runtime new memory manager which adds memory pools allocations.

* address comments.

---------

Co-authored-by: Venkateshwar Reddy Kandula <vkandula@amd.com>

[ROCm/rocprofiler-sdk commit: 176f8aea67]
2025-04-24 19:55:45 -05:00
Kandula, Venkateshwar reddy 176f8aea67 [PSDB] [CI] SWDEV-528922: Modify summary MEMORY_ALLOCATION kind tests to account new hip runtime memory manger (#368)
* Modify summary MEMORY_ALLOCATION kind tests to account new hip runtime new memory manager which adds memory pools allocations.

* address comments.

---------

Co-authored-by: Venkateshwar Reddy Kandula <vkandula@amd.com>
2025-04-24 19:55:45 -05:00
Li, Todd tiantuo 8706df3726 SWDEV-511055 - fix HIP PAL memory allocation workaround for APU (#40)
[ROCm/clr commit: 95cdc83eaf]
2025-04-24 15:07:16 -07:00
Li, Todd tiantuo 95cdc83eaf SWDEV-511055 - fix HIP PAL memory allocation workaround for APU (#40) 2025-04-24 15:07:16 -07:00
Sang, Tao a9068182f4 SWDEV-493275 - Support scratch limit (#20)
Support programmatic query and change of scratch limit on
AMD devices.

Change-Id: Id5da355a77366f97868e462847f3916e87fd2af6

[ROCm/clr commit: 1113eff3f9]
2025-04-24 17:15:25 -04:00
Sang, Tao 1113eff3f9 SWDEV-493275 - Support scratch limit (#20)
Support programmatic query and change of scratch limit on
AMD devices.

Change-Id: Id5da355a77366f97868e462847f3916e87fd2af6
2025-04-24 17:15:25 -04:00
Critchley, Paul 4fd8a0164a SWDEV-527731 - [Ubertrace] OpenCL driver reports wrong Instrumentation API Version (#211)
[ROCm/clr commit: 4f2a4b12a9]
2025-04-24 14:06:17 -07:00
Critchley, Paul 4f2a4b12a9 SWDEV-527731 - [Ubertrace] OpenCL driver reports wrong Instrumentation API Version (#211) 2025-04-24 14:06:17 -07:00
Pratik Basyal 9f0ea55807 Changelog edited to sync with release branch and clarity [amdstaging] (#182)
* Changelog edited to sync with release-6.4 branch

* Peer review feedback added

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

---------
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

[ROCm/rocprofiler-systems commit: 8b09fb71d9]
2025-04-24 16:19:13 -04:00
Pratik Basyal 8b09fb71d9 Changelog edited to sync with release branch and clarity [amdstaging] (#182)
* Changelog edited to sync with release-6.4 branch

* Peer review feedback added

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

---------
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
2025-04-24 16:19:13 -04:00
Pratik Basyal 6b93df5843 Broken link fix to How-to: Configuring Runtime Options [amd-staging] (#180)
* Update to KOKKOS_TOOLS_LIBS env var (#69)

* Broken link fixed

---------

Co-authored-by: Maarten Arnst <maarten.arnst@ulg.ac.be>

[ROCm/rocprofiler-systems commit: 16745948de]
2025-04-24 12:34:00 -04:00
Pratik Basyal 16745948de Broken link fix to How-to: Configuring Runtime Options [amd-staging] (#180)
* Update to KOKKOS_TOOLS_LIBS env var (#69)

* Broken link fixed

---------

Co-authored-by: Maarten Arnst <maarten.arnst@ulg.ac.be>
2025-04-24 12:34:00 -04:00
Bhardwaj, Gopesh 20bc0deaf1 doc improvements for 1.0.0 (#367)
* correcting rocprofiler_configure

* Fix indexing order

* doc feedback

[ROCm/rocprofiler-sdk commit: bbe9eab53a]
2025-04-24 17:05:22 +05:30
Bhardwaj, Gopesh bbe9eab53a doc improvements for 1.0.0 (#367)
* correcting rocprofiler_configure

* Fix indexing order

* doc feedback
2025-04-24 17:05:22 +05:30
Godavarthy Surya, Anusha d4e36d0900 SWDEV-469423 - hipStreamEndCapture graph* can be nullptr (#170)
[ROCm/clr commit: e5ce544c45]
2025-04-24 13:57:09 +05:30
Godavarthy Surya, Anusha e5ce544c45 SWDEV-469423 - hipStreamEndCapture graph* can be nullptr (#170) 2025-04-24 13:57:09 +05:30
Hernandez, Gerardo 817eb2a3d5 Disable deviceAllocation tests (#141)
Disable tests calling TestMemoryAcrossMulKernels()/TestMemoryAcrossMulKernelsUsingGraph() as they might be hanging the whole machine on Windows

[ROCm/hip-tests commit: a423c20690]
2025-04-24 13:39:04 +05:30
Hernandez, Gerardo a423c20690 Disable deviceAllocation tests (#141)
Disable tests calling TestMemoryAcrossMulKernels()/TestMemoryAcrossMulKernelsUsingGraph() as they might be hanging the whole machine on Windows
2025-04-24 13:39:04 +05:30