76333 Commits

Author SHA1 Message Date
Madsen, Jonathan 96ec52f2da rocprofv3: do not abort if counter does not have dimensions (#150)
* rocprofv3: do not abort if counter does not have dimensions

* Relax error handling further in rocprofv3 metadata

---------

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

[ROCm/rocprofiler-sdk commit: 3071199386]
2025-02-11 14:31:25 -06:00
Madsen, Jonathan 3071199386 rocprofv3: do not abort if counter does not have dimensions (#150)
* rocprofv3: do not abort if counter does not have dimensions

* Relax error handling further in rocprofv3 metadata

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
2025-02-11 14:31:25 -06:00
Wenkai Du 7fdbcdfdec Move collective trace to HBM and fix log issue (#1542)
[ROCm/rccl commit: f5b15f27a9]
2025-02-11 11:40:14 -08:00
Wenkai Du f5b15f27a9 Move collective trace to HBM and fix log issue (#1542) 2025-02-11 11:40:14 -08:00
Kandula, Venkateshwar reddy a585468121 [BUG FIX] store dimensions in counter id when used reduce operator (#181)
* save other dimension in counter id.

* Formating

---------

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

[ROCm/rocprofiler-sdk commit: 143f84fe6b]
2025-02-11 13:05:57 -06:00
Kandula, Venkateshwar reddy 143f84fe6b [BUG FIX] store dimensions in counter id when used reduce operator (#181)
* save other dimension in counter id.

* Formating

---------

Co-authored-by: Venkateshwar Reddy Kandula <vkandula@amd.com>
2025-02-11 13:05:57 -06:00
Rahul Manocha 6cde56faa8 SWDEV-492165 - Add new test types for atomicCAS and unsafeAtomicAdd
Change-Id: Icacb6e94c64c5d4ce3f15c6d4a50fe8506cedfa9


[ROCm/hip-tests commit: c7236e3e2e]
2025-02-11 13:21:22 -05:00
Rahul Manocha c7236e3e2e SWDEV-492165 - Add new test types for atomicCAS and unsafeAtomicAdd
Change-Id: Icacb6e94c64c5d4ce3f15c6d4a50fe8506cedfa9
2025-02-11 13:21:22 -05:00
Sajina PK fba64f8acd Add support for VA-API and rocDecode tracing (#92)
- VA API tracing using Timemory gotcha wrappers.
- rocDecode API tracing integration using callback to ROCPROFILER_CALLBACK_TRACING_ROCDECODE_API
- Updated videodecode ctest to validate rocDecode APIs in perfetto trace. 

[ROCm/rocprofiler-systems commit: 697d1ac02f]
2025-02-11 13:08:23 -05:00
Sajina PK 697d1ac02f Add support for VA-API and rocDecode tracing (#92)
- VA API tracing using Timemory gotcha wrappers.
- rocDecode API tracing integration using callback to ROCPROFILER_CALLBACK_TRACING_ROCDECODE_API
- Updated videodecode ctest to validate rocDecode APIs in perfetto trace.
2025-02-11 13:08:23 -05:00
rahulc1984 689725fb9e Make rccl version detection robust. (#1517)
* Accept an EXPLICIT_ROCM_VERSION and use that vs inspecting the environment if provided.
* Use CMake's built in file reading support vs execute_process (without error checking) to avoid silent but deadly later failures.
* Properly quote some comparisons to avoid syntax errors if they happen to have an empty string.
* Guard against ROCM_PATH being an empty string, avoiding stray path extensions to root directories, etc.

Co-authored-by: Stella Laurenzo <stellaraccident@gmail.com>

[ROCm/rccl commit: 92ac136db5]
2025-02-11 10:48:22 -07:00
rahulc1984 92ac136db5 Make rccl version detection robust. (#1517)
* Accept an EXPLICIT_ROCM_VERSION and use that vs inspecting the environment if provided.
* Use CMake's built in file reading support vs execute_process (without error checking) to avoid silent but deadly later failures.
* Properly quote some comparisons to avoid syntax errors if they happen to have an empty string.
* Guard against ROCM_PATH being an empty string, avoiding stray path extensions to root directories, etc.

Co-authored-by: Stella Laurenzo <stellaraccident@gmail.com>
2025-02-11 10:48:22 -07:00
cfallows-amd 611704d825 Update CHANGELOG and docs
[ROCm/rocprofiler-compute commit: c03cbf2389]
2025-02-11 10:46:33 -07:00
cfallows-amd c03cbf2389 Update CHANGELOG and docs 2025-02-11 10:46:33 -07:00
Lakshmi Kumar 2d65d06e92 gfx update (#118)
[ROCm/rocjpeg commit: 0442180916]
2025-02-11 12:42:34 -05:00
Lakshmi Kumar 0442180916 gfx update (#118) 2025-02-11 12:42:34 -05:00
Madsen, Jonathan d7495f9f1a Re-enable clang-tidy for core workflows + clang-tidy fixes (#197)
* Ensure the clang-tidy is updated + clang-tidy fixes

* update-ci workflow

* Enable clang-tidy checks

* Add extra logging to device counter collection samples

* Misc clang-tidy fixes

* Disable device counter collection samples for ThreadSanitizer

* Formatting

---------

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

[ROCm/rocprofiler-sdk commit: 070b659a9a]
2025-02-11 10:58:47 -06:00
Madsen, Jonathan 070b659a9a Re-enable clang-tidy for core workflows + clang-tidy fixes (#197)
* Ensure the clang-tidy is updated + clang-tidy fixes

* update-ci workflow

* Enable clang-tidy checks

* Add extra logging to device counter collection samples

* Misc clang-tidy fixes

* Disable device counter collection samples for ThreadSanitizer

* Formatting

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
2025-02-11 10:58:47 -06:00
corey-derochie-amd 30e7047750 Switched from cmake_host_system_information feature to a manual parse (#1518)
* Switched cmake_host_system_information feature to a manual parse to remain cmake 3.5 compliant.

* Updating minimum cmake to 3.16 to conform with the rest of ROCm. This change still applies.

[ROCm/rccl commit: 42ab425037]
2025-02-11 08:51:39 -07:00
corey-derochie-amd 42ab425037 Switched from cmake_host_system_information feature to a manual parse (#1518)
* Switched cmake_host_system_information feature to a manual parse to remain cmake 3.5 compliant.

* Updating minimum cmake to 3.16 to conform with the rest of ROCm. This change still applies.
2025-02-11 08:51:39 -07:00
Nilesh M Negi 4ccbaabdc9 [UT] Include iomanip if not defined (#1510)
* [UT] Include iomanip if not defined

Signed-off-by: nileshnegi <Nilesh.Negi@amd.com>

* Remove include guards

`iomanip.h` has pre-defined include guards. These are not needed.

Co-authored-by: corey-derochie-amd <161367113+corey-derochie-amd@users.noreply.github.com>

---------

Signed-off-by: nileshnegi <Nilesh.Negi@amd.com>
Co-authored-by: corey-derochie-amd <161367113+corey-derochie-amd@users.noreply.github.com>

[ROCm/rccl commit: 4e406acc43]
2025-02-11 08:48:47 -07:00
Nilesh M Negi 4e406acc43 [UT] Include iomanip if not defined (#1510)
* [UT] Include iomanip if not defined

Signed-off-by: nileshnegi <Nilesh.Negi@amd.com>

* Remove include guards

`iomanip.h` has pre-defined include guards. These are not needed.

Co-authored-by: corey-derochie-amd <161367113+corey-derochie-amd@users.noreply.github.com>

---------

Signed-off-by: nileshnegi <Nilesh.Negi@amd.com>
Co-authored-by: corey-derochie-amd <161367113+corey-derochie-amd@users.noreply.github.com>
2025-02-11 08:48:47 -07:00
Lakshmi Kumar 348b73425b gfx update (#505)
Co-authored-by: Kiriti Gowda <kiritigowda@gmail.com>

[ROCm/rocdecode commit: 5e806d5465]
2025-02-11 09:07:34 -05:00
Lakshmi Kumar 5e806d5465 gfx update (#505)
Co-authored-by: Kiriti Gowda <kiritigowda@gmail.com>
2025-02-11 09:07:34 -05:00
Liu, Shuzhou (Bill) 8e74024b11 Change default mutex to thread only (#104)
Most sysfs reads do not require cross-process level mutex, and writes to sysfs
should be protected by the kernel already.
Users can still switch to the old behavior by setting the environment variable AMDSMI_MUTEX_CROSS_PROCESS=1.
2025-02-11 07:27:03 -06:00
Liu, Shuzhou (Bill) 4f750bd4d5 Change default mutex to thread only (#104)
Most sysfs reads do not require cross-process level mutex, and writes to sysfs
should be protected by the kernel already.
Users can still switch to the old behavior by setting the environment variable AMDSMI_MUTEX_CROSS_PROCESS=1.


[ROCm/amdsmi commit: 8e74024b11]
2025-02-11 07:27:03 -06:00
kjayapra-amd 1f648c7d94 SWDEV-511672 - Special case the Remote USWC memory usage for HIP, if the alloc size is large.
Change-Id: I524c1402b249cedfd58b56f494caa2ac057e1623


[ROCm/clr commit: cf6aabb823]
2025-02-11 06:42:18 -05:00
kjayapra-amd cf6aabb823 SWDEV-511672 - Special case the Remote USWC memory usage for HIP, if the alloc size is large.
Change-Id: I524c1402b249cedfd58b56f494caa2ac057e1623
2025-02-11 06:42:18 -05:00
Arif, Maisam 552f15a1fb Fixed RDC to work with updated amdsmi_get_power_info() (#115)
Change-Id: Ic9e7a68ae58f61dbe73fc7d1b17af34152933e71

Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
2025-02-11 00:51:29 -06:00
Arif, Maisam c26abbbe9a Fixed RDC to work with updated amdsmi_get_power_info() (#115)
Change-Id: Ic9e7a68ae58f61dbe73fc7d1b17af34152933e71

Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>

[ROCm/rdc commit: 552f15a1fb]
2025-02-11 00:51:29 -06:00
Saleel Kudchadker 71e1a0b10d SWDEV-504494 - Further copy improvements
- Fix regression for D2H pinned copies which adds systemscope release.
- Skip cpu wait for D2H unpinned copies as we can pass the signal of the
  barrier to rocr copy.
- Fix an old bug in sdmaEngineRetainCount_ logic
- Improve logging

Change-Id: If074bddb05564b15949b0d5f9bf12acd3692174e


[ROCm/clr commit: 4c95ee5e1e]
2025-02-11 00:55:52 -05:00
Saleel Kudchadker 4c95ee5e1e SWDEV-504494 - Further copy improvements
- Fix regression for D2H pinned copies which adds systemscope release.
- Skip cpu wait for D2H unpinned copies as we can pass the signal of the
  barrier to rocr copy.
- Fix an old bug in sdmaEngineRetainCount_ logic
- Improve logging

Change-Id: If074bddb05564b15949b0d5f9bf12acd3692174e
2025-02-11 00:55:52 -05:00
Bhardwaj, Gopesh 6eb343aa4a Adding pc sampling how to guide (#160)
* Adding pc sampling how to guide

* doc update

* Fixing indentation

* updating index

* udpating doc

* updating doc

* Added field information

* Fixing Formatting

* fix formatting error

* Added json format for pc sampling

* feedback resolved

* formatting for text

* PC Sampling API doc

* Reformatted

* Note for shared systems

* update docs

* correcting relative path for cross-referencing

---------

Co-authored-by: vlaindic_amdeng <vladimir.indic@amd.com>

[ROCm/rocprofiler-sdk commit: cdf22eba7d]
2025-02-10 20:33:05 -06:00
Bhardwaj, Gopesh cdf22eba7d Adding pc sampling how to guide (#160)
* Adding pc sampling how to guide

* doc update

* Fixing indentation

* updating index

* udpating doc

* updating doc

* Added field information

* Fixing Formatting

* fix formatting error

* Added json format for pc sampling

* feedback resolved

* formatting for text

* PC Sampling API doc

* Reformatted

* Note for shared systems

* update docs

* correcting relative path for cross-referencing

---------

Co-authored-by: vlaindic_amdeng <vladimir.indic@amd.com>
2025-02-10 20:33:05 -06:00
Elwazir, Ammar f8bff7b835 Disabling Mi325 temp. (#199)
[ROCm/rocprofiler-sdk commit: c478c24616]
2025-02-10 20:07:20 -06:00
Elwazir, Ammar c478c24616 Disabling Mi325 temp. (#199) 2025-02-10 20:07:20 -06:00
ywang103-amd 5cc49d6e90 fix crash for running rocprofv3 on mi100 (#557)
* initial hack to fix for v3 stucking on mi100 becasue of -m parameter and missing counter csv file

* proper formating

* refactored profiler option function to take soc arch

* resolve missing step that casued error for profiler option

* fix typo of arch name

* change method of putting soc info into profiler option

* isort and black format

* add comment for the part that handles missing counter csv file

* remove unncecessary import

---------

Co-authored-by: YANG WANG <ywang@ywang-ubuntu.amd.com>

[ROCm/rocprofiler-compute commit: 5ee37b3353]
2025-02-10 17:59:03 -05:00
ywang103-amd 5ee37b3353 fix crash for running rocprofv3 on mi100 (#557)
* initial hack to fix for v3 stucking on mi100 becasue of -m parameter and missing counter csv file

* proper formating

* refactored profiler option function to take soc arch

* resolve missing step that casued error for profiler option

* fix typo of arch name

* change method of putting soc info into profiler option

* isort and black format

* add comment for the part that handles missing counter csv file

* remove unncecessary import

---------

Co-authored-by: YANG WANG <ywang@ywang-ubuntu.amd.com>
2025-02-10 17:59:03 -05:00
xuchen-amd 47921fe78d Change profiler mode trace options, the different trace modes (kernel, kokkos (marker), and hip) are now exclusive of each other based on rocprofv3 updates. (#561)
[ROCm/rocprofiler-compute commit: 6391b3ca76]
2025-02-10 16:11:03 -05:00
xuchen-amd 6391b3ca76 Change profiler mode trace options, the different trace modes (kernel, kokkos (marker), and hip) are now exclusive of each other based on rocprofv3 updates. (#561) 2025-02-10 16:11:03 -05:00
Harish Kasiviswanathan 2a64fa5e06 libhsakmt: gfx950: Add option to enable HIGH_PRECISION
Environment variable HSA_HIGH_PRECISION_MODE can be used to control MFMA
precision

Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Change-Id: Ib78dd9dd8867025e090a3cca96ab6db4f65dea12
2025-02-10 16:05:25 -05:00
Harish Kasiviswanathan 729f98b05f libhsakmt: gfx950: Add option to enable HIGH_PRECISION
Environment variable HSA_HIGH_PRECISION_MODE can be used to control MFMA
precision

Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Change-Id: Ib78dd9dd8867025e090a3cca96ab6db4f65dea12


[ROCm/ROCR-Runtime commit: 2a64fa5e06]
2025-02-10 16:05:25 -05:00
Yiltan 1f3881fa21 Merge pull request #38 from Yiltan/ro/implement-sigops
Implements Signalling Operations for RO

[ROCm/rocshmem commit: 495cd6970b]
2025-02-10 15:10:07 -05:00
Yiltan 495cd6970b Merge pull request #38 from Yiltan/ro/implement-sigops
Implements Signalling Operations for RO
2025-02-10 15:10:07 -05:00
Yiltan 87e049f9c9 Merge pull request #34 from Yiltan/sigops-test-fix
Fix Signalling Operations Functional Test

[ROCm/rocshmem commit: 94144f4460]
2025-02-10 14:56:45 -05:00
Yiltan 94144f4460 Merge pull request #34 from Yiltan/sigops-test-fix
Fix Signalling Operations Functional Test
2025-02-10 14:56:45 -05:00
Yiltan f64c76b31c Merge pull request #39 from Yiltan/ro/fix-teamreduce
Fix Team reduction intra-node

[ROCm/rocshmem commit: 944444cf12]
2025-02-10 14:56:27 -05:00
Yiltan 944444cf12 Merge pull request #39 from Yiltan/ro/fix-teamreduce
Fix Team reduction intra-node
2025-02-10 14:56:27 -05:00
Welton, Benjamin b90f127957 [SWDEV-513658] Force HSA_AMD_MEMORY_POOL_EXECUTABLE_FLAG value to be used with HSA calls (#192)
* Force HSA_AMD_MEMORY_POOL_EXECUTABLE_FLAG  value to be used with HSA calls

Fix for CI

* More tweaks

* Increase reproducible-runtime kernel sleep granularity

* Fix data race in synchronous device counter collection sample

* Update device counting service

- add get_active_context function

---------

Co-authored-by: Benjamin Welton <bewelton@amd.com>
Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>

[ROCm/rocprofiler-sdk commit: 080b2ba451]
2025-02-10 11:34:26 -06:00
Welton, Benjamin 080b2ba451 [SWDEV-513658] Force HSA_AMD_MEMORY_POOL_EXECUTABLE_FLAG value to be used with HSA calls (#192)
* Force HSA_AMD_MEMORY_POOL_EXECUTABLE_FLAG  value to be used with HSA calls

Fix for CI

* More tweaks

* Increase reproducible-runtime kernel sleep granularity

* Fix data race in synchronous device counter collection sample

* Update device counting service

- add get_active_context function

---------

Co-authored-by: Benjamin Welton <bewelton@amd.com>
Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
2025-02-10 11:34:26 -06:00