Граф коммитов

64512 Коммитов

Автор SHA1 Сообщение Дата
Anusha GodavarthySurya 0bbe324c18 SWDEV-448586 - Added new API hipStreamBeginCaptureToGraph
Change-Id: I0e101bcdc09e3770682df07b1ef6c07c4669ea49


[ROCm/hip commit: edf2e473e3]
2024-03-01 03:43:59 -05:00
Anusha GodavarthySurya edf2e473e3 SWDEV-448586 - Added new API hipStreamBeginCaptureToGraph
Change-Id: I0e101bcdc09e3770682df07b1ef6c07c4669ea49
2024-03-01 03:43:59 -05:00
Jonathan R. Madsen cb6b79c323 Fix rocprofiler_iterate_callback_tracing_kind_operation_args for HIP compiler callbacks (#532)
* Fix HIP compiler iterate args

- `include/rocprofiler-sdk/hip/api_args.h`
  - replace struct fields named "f" with "func"
  - replace hip stream fields named "hStream" with "stream"
- `lib/rocprofiler-sdk/callback_tracing.cpp`
  - iterate_args for HIP compiler table
- `lib/rocprofiler-sdk/registration.cpp`
  - fix warning about roctx num_tables
- `lib/rocprofiler-sdk/hip/hip.def.cpp`
  - replace struct fields named "f" with "func"
  - replace hip stream fields named "hStream" with "stream"
- `lib/rocprofiler-sdk/{hip,hsa,marker}/utils.hpp`
  - improve `stringize_impl`
- `lib/rocprofiler-sdk/hsa/code_object.cpp`
  - remove stale commented out code
- `lib/rocprofiler-sdk/hsa/queue_controller.*`
  - destory_queue -> destroy_queue
- `tests/tools/json-tool.cpp`
  - improve parallelism in tool_tracing_callback
  - serialize the marker api args
  - only invoke rocprofiler_iterate_callback_tracing_kind_operation_args in exit phase
- `samples/counter_collection/CMakeLists.txt`
  - reduce timeout on tests to 120 seconds

* Update lib/rocprofiler-sdk/hsa/utils.hpp

- disable dereference of double pointer in stringize_impl

* Update lib/common

- indirection_level in mpl.hpp
- stringize_arg.hpp

* Rework rocprofiler_iterate_callback_tracing_kind_operation_args

- provide more information in rocprofiler_callback_tracing_operation_args_cb_t
- support specifying the dereference level to account for output paramters

[ROCm/rocprofiler-sdk commit: 1bb94add11]
2024-03-01 01:46:07 -06:00
Jonathan R. Madsen 1bb94add11 Fix rocprofiler_iterate_callback_tracing_kind_operation_args for HIP compiler callbacks (#532)
* Fix HIP compiler iterate args

- `include/rocprofiler-sdk/hip/api_args.h`
  - replace struct fields named "f" with "func"
  - replace hip stream fields named "hStream" with "stream"
- `lib/rocprofiler-sdk/callback_tracing.cpp`
  - iterate_args for HIP compiler table
- `lib/rocprofiler-sdk/registration.cpp`
  - fix warning about roctx num_tables
- `lib/rocprofiler-sdk/hip/hip.def.cpp`
  - replace struct fields named "f" with "func"
  - replace hip stream fields named "hStream" with "stream"
- `lib/rocprofiler-sdk/{hip,hsa,marker}/utils.hpp`
  - improve `stringize_impl`
- `lib/rocprofiler-sdk/hsa/code_object.cpp`
  - remove stale commented out code
- `lib/rocprofiler-sdk/hsa/queue_controller.*`
  - destory_queue -> destroy_queue
- `tests/tools/json-tool.cpp`
  - improve parallelism in tool_tracing_callback
  - serialize the marker api args
  - only invoke rocprofiler_iterate_callback_tracing_kind_operation_args in exit phase
- `samples/counter_collection/CMakeLists.txt`
  - reduce timeout on tests to 120 seconds

* Update lib/rocprofiler-sdk/hsa/utils.hpp

- disable dereference of double pointer in stringize_impl

* Update lib/common

- indirection_level in mpl.hpp
- stringize_arg.hpp

* Rework rocprofiler_iterate_callback_tracing_kind_operation_args

- provide more information in rocprofiler_callback_tracing_operation_args_cb_t
- support specifying the dereference level to account for output paramters
2024-03-01 01:46:07 -06:00
ChingShihLi 5ca2131ce0 SWDEV-413377 - Segfault for hipLaunchKernel with multi GPUs
- Using runtime unbundler, no any gfx device can load fat binary,
if there is any device without available code object.

- Extract available code object to corresponding gfx devices. So
users can work ROCm with those ready devices without segmentation
fault.

Change-Id: I9f14c65ecebf2d3c4b127a007cb434a3ae98c450


[ROCm/clr commit: 6723277ad4]
2024-03-01 02:24:22 -05:00
ChingShihLi 6723277ad4 SWDEV-413377 - Segfault for hipLaunchKernel with multi GPUs
- Using runtime unbundler, no any gfx device can load fat binary,
if there is any device without available code object.

- Extract available code object to corresponding gfx devices. So
users can work ROCm with those ready devices without segmentation
fault.

Change-Id: I9f14c65ecebf2d3c4b127a007cb434a3ae98c450
2024-03-01 02:24:22 -05:00
Jonathan R. Madsen 15302ff11d C compatibility for public headers (#566)
* C compatibility for public headers

- add tests/tools/c-tool.c
  - builds a tool (which does nothing) with C language
  - ensures that tool can be compiled in C
- add tests/c-tool/CMakeLists.txt
  - ensures that tool library build from C is a valid tool
- rocprofiler_counter_info_v0_t is_derived is int instead of bool
  - C does not have bool unless <stdbool.h> is included
- add `include/rocprofiler-sdk/hsa/api_trace_version.h
  - handles providing HSA_*_TABLE_(MAJOR|STEP)_VERSION values if compiled from C
- cmake define in version.h.in for ROCPROFILER_HSA_*_TABLE_(MAJOR|STEP)_VERSION
  - HSA table versions compiled with
- use rocprofiler_(hsa|hip|marker)_api_no_args struct to handle incompatibility b/t empty structs in C vs. C++ (size of 0 vs. size of 1)
- extern "C" in include/rocprofiler-sdk/{hsa,hip,marker}/api_args.h
- fixed spelling error: derrived -> derived
- scope YY_NO_INPUT compile definition to lib/rocprofiler-sdk/counters/parser/*

* Revert CDash dashboard

[ROCm/rocprofiler-sdk commit: a1267e1fd2]
2024-02-29 23:49:54 -06:00
Jonathan R. Madsen a1267e1fd2 C compatibility for public headers (#566)
* C compatibility for public headers

- add tests/tools/c-tool.c
  - builds a tool (which does nothing) with C language
  - ensures that tool can be compiled in C
- add tests/c-tool/CMakeLists.txt
  - ensures that tool library build from C is a valid tool
- rocprofiler_counter_info_v0_t is_derived is int instead of bool
  - C does not have bool unless <stdbool.h> is included
- add `include/rocprofiler-sdk/hsa/api_trace_version.h
  - handles providing HSA_*_TABLE_(MAJOR|STEP)_VERSION values if compiled from C
- cmake define in version.h.in for ROCPROFILER_HSA_*_TABLE_(MAJOR|STEP)_VERSION
  - HSA table versions compiled with
- use rocprofiler_(hsa|hip|marker)_api_no_args struct to handle incompatibility b/t empty structs in C vs. C++ (size of 0 vs. size of 1)
- extern "C" in include/rocprofiler-sdk/{hsa,hip,marker}/api_args.h
- fixed spelling error: derrived -> derived
- scope YY_NO_INPUT compile definition to lib/rocprofiler-sdk/counters/parser/*

* Revert CDash dashboard
2024-02-29 23:49:54 -06:00
jujiang 3cde9aacb2 SWDEV-444835 - Porting changelog from 6.1 to staging
Change-Id: I85b2993485da797df4ac3e146481c341680c4595


[ROCm/clr commit: 574d946cfb]
2024-02-29 23:44:02 -05:00
jujiang 574d946cfb SWDEV-444835 - Porting changelog from 6.1 to staging
Change-Id: I85b2993485da797df4ac3e146481c341680c4595
2024-02-29 23:44:02 -05:00
Saleel Kudchadker 939642fed7 SWDEV-301667 - Increase default signal pool to 4096
Change-Id: I4ab23b0f87e295b40ab76ad6e96249d11b8ad04d


[ROCm/clr commit: 94c7004df8]
2024-02-29 22:52:02 +00:00
Saleel Kudchadker 94c7004df8 SWDEV-301667 - Increase default signal pool to 4096
Change-Id: I4ab23b0f87e295b40ab76ad6e96249d11b8ad04d
2024-02-29 22:52:02 +00:00
Giovanni LB c6575ab01c SWDEV-448279: Disabling CTF tests
Change-Id: I63f915862fe28b51206ed58424d3358385103110
2024-02-29 18:43:50 -03:00
Giovanni LB 36377e3d51 SWDEV-448279: Disabling CTF tests
Change-Id: I63f915862fe28b51206ed58424d3358385103110


[ROCm/rocprofiler commit: c6575ab01c]
2024-02-29 18:43:50 -03:00
German Andryeyev 01c1d8f568 SWDEV-311271 - Fix mempool tests
- Reduce the number of blocks in execution, because small allocations
can cause page faults and make the system unstable after
- Change error codes

Change-Id: I77561ffa236e05ad08afdcc79b7770bf36fcdf50


[ROCm/hip-tests commit: 8ef86d1c7c]
2024-02-29 15:21:20 -05:00
German Andryeyev 8ef86d1c7c SWDEV-311271 - Fix mempool tests
- Reduce the number of blocks in execution, because small allocations
can cause page faults and make the system unstable after
- Change error codes

Change-Id: I77561ffa236e05ad08afdcc79b7770bf36fcdf50
2024-02-29 15:21:20 -05:00
JoseSantosAMD f240822ea1 Fix off by one out of range kernel filtering (#287)
* Fi off by one out of range kernel filtering

Signed-off-by: JoseSantosAMD <Jose.Santos@amd.com>

* reformat

Signed-off-by: JoseSantosAMD <Jose.Santos@amd.com>

---------

Signed-off-by: JoseSantosAMD <Jose.Santos@amd.com>

[ROCm/rocprofiler-compute commit: 4ad124a207]
2024-02-29 12:38:37 -06:00
JoseSantosAMD 4ad124a207 Fix off by one out of range kernel filtering (#287)
* Fi off by one out of range kernel filtering

Signed-off-by: JoseSantosAMD <Jose.Santos@amd.com>

* reformat

Signed-off-by: JoseSantosAMD <Jose.Santos@amd.com>

---------

Signed-off-by: JoseSantosAMD <Jose.Santos@amd.com>
2024-02-29 12:38:37 -06:00
JoseSantosAMD 055bda677a Supress warning: RuntimeWarning: Mean of empty slice (#271)
* Supress warning: RuntimeWarning: Mean of empty slice

Signed-off-by: Jose Santos <josantos@login1.hpcfund>

* Move import to top

Signed-off-by: Jose Santos <josantos@login1.hpcfund>

* Specify type of RuntimeWarning in catcher. Safer approach

Signed-off-by: colramos-amd <colramos@amd.com>

* Comply to formatting

Signed-off-by: colramos-amd <colramos@amd.com>

---------

Signed-off-by: Jose Santos <josantos@login1.hpcfund>
Signed-off-by: colramos-amd <colramos@amd.com>
Co-authored-by: Jose Santos <josantos@login1.hpcfund>
Co-authored-by: colramos-amd <colramos@amd.com>

[ROCm/rocprofiler-compute commit: f42aa6c7e0]
2024-02-29 12:14:46 -06:00
JoseSantosAMD f42aa6c7e0 Supress warning: RuntimeWarning: Mean of empty slice (#271)
* Supress warning: RuntimeWarning: Mean of empty slice

Signed-off-by: Jose Santos <josantos@login1.hpcfund>

* Move import to top

Signed-off-by: Jose Santos <josantos@login1.hpcfund>

* Specify type of RuntimeWarning in catcher. Safer approach

Signed-off-by: colramos-amd <colramos@amd.com>

* Comply to formatting

Signed-off-by: colramos-amd <colramos@amd.com>

---------

Signed-off-by: Jose Santos <josantos@login1.hpcfund>
Signed-off-by: colramos-amd <colramos@amd.com>
Co-authored-by: Jose Santos <josantos@login1.hpcfund>
Co-authored-by: colramos-amd <colramos@amd.com>
2024-02-29 12:14:46 -06:00
Karl W Schulz 0dc03e2157 Update CI configuration
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: e63c24c618]
2024-02-29 10:58:38 -06:00
Karl W Schulz e63c24c618 Update CI configuration
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
2024-02-29 10:58:38 -06:00
jujiang 855a59c16c SWDEV-444835 - Merge hip-tests changes made in 6.1- to staging
Change-Id: Ib0e926467dce61d3b8ca3bd9e8f0b682b666b0c7


[ROCm/hip-tests commit: 4a2fbe2e1f]
2024-02-29 11:38:04 -05:00
jujiang 4a2fbe2e1f SWDEV-444835 - Merge hip-tests changes made in 6.1- to staging
Change-Id: Ib0e926467dce61d3b8ca3bd9e8f0b682b666b0c7
2024-02-29 11:38:04 -05:00
Rakesh Roy 468b69ed94 SWDEV-446588 - Disable graph multi gpu testcases
- Disable graph multi gpu testcases until graph has support for it

Change-Id: I7be60247f649024f7468c71e374698d0fab46c3f


[ROCm/hip-tests commit: f7347f5ea2]
2024-02-29 11:04:17 +05:30
Rakesh Roy f7347f5ea2 SWDEV-446588 - Disable graph multi gpu testcases
- Disable graph multi gpu testcases until graph has support for it

Change-Id: I7be60247f649024f7468c71e374698d0fab46c3f
2024-02-29 11:04:17 +05:30
Giovanni LB c53b6ff6f7 SWDEV-448279: Workaround for UB24 compiler errors
Change-Id: I01e880502edac8e8b941d199e08bc8f1eded0a89


[ROCm/roctracer commit: 4bc8dbd029]
2024-02-28 23:56:14 -03:00
Giovanni LB 4bc8dbd029 SWDEV-448279: Workaround for UB24 compiler errors
Change-Id: I01e880502edac8e8b941d199e08bc8f1eded0a89
2024-02-28 23:56:14 -03:00
Saleel Kudchadker ce7b62d15c SWDEV-443760 - Enable device kernel args for MI300
- Enable Device kernel args for MI300* for now.
- Fix a perf issue which impacts graph instantiate when dev kernel args
are enabled.

Change-Id: I962e58fd9d8dd1a8db95e601cb03a8e9c7bac97f


[ROCm/clr commit: 68f40f78dd]
2024-02-28 19:10:04 -05:00
Saleel Kudchadker 68f40f78dd SWDEV-443760 - Enable device kernel args for MI300
- Enable Device kernel args for MI300* for now.
- Fix a perf issue which impacts graph instantiate when dev kernel args
are enabled.

Change-Id: I962e58fd9d8dd1a8db95e601cb03a8e9c7bac97f
2024-02-28 19:10:04 -05:00
Rahul Manocha 0a269e98a9 [SWDEV-448077][SWDEV-448067] - Changes to Kernel Node Attribute related APIs
Change-Id: Ibbf773fd5f134a62b7ce04f6956b10c1086b1782


[ROCm/clr commit: 2c0fa829b4]
2024-02-28 18:14:41 -05:00
Rahul Manocha 2c0fa829b4 [SWDEV-448077][SWDEV-448067] - Changes to Kernel Node Attribute related APIs
Change-Id: Ibbf773fd5f134a62b7ce04f6956b10c1086b1782
2024-02-28 18:14:41 -05:00
Gopesh Bhardwaj 495ad364a2 counter collection multi-kernel test (#492)
* counter collection multi-kernel test

* Adding counter collection in rocprofv3 script

[ROCm/rocprofiler-sdk commit: 31dcfabe23]
2024-02-28 17:08:28 -06:00
Gopesh Bhardwaj 31dcfabe23 counter collection multi-kernel test (#492)
* counter collection multi-kernel test

* Adding counter collection in rocprofv3 script
2024-02-28 17:08:28 -06:00
Anusha GodavarthySurya f1e5162757 SWDEV-447545 - Fix Enable/Disable node with hipGraph
Node can be enabled/disabled only for kernel, memcpy and memset nodes.
If the node is disabled it becomes empty node.
To maintain ordering just enqueue marker with respective node dependencies.

Change-Id: I710f3e88ab4e76c81f6f86a40a7dc61fd4c7e440


[ROCm/clr commit: e0e63eb04d]
2024-02-28 17:34:03 -05:00
Anusha GodavarthySurya e0e63eb04d SWDEV-447545 - Fix Enable/Disable node with hipGraph
Node can be enabled/disabled only for kernel, memcpy and memset nodes.
If the node is disabled it becomes empty node.
To maintain ordering just enqueue marker with respective node dependencies.

Change-Id: I710f3e88ab4e76c81f6f86a40a7dc61fd4c7e440
2024-02-28 17:34:03 -05:00
Anusha GodavarthySurya 67543a60f1 SWDEV-444988 - Fix hiddenheap variable initialization
Change-Id: Icdb32ef4718b8ca0beedd569a4aaf40a9f4fb49d


[ROCm/clr commit: ea66946399]
2024-02-28 17:22:32 -05:00
Anusha GodavarthySurya ea66946399 SWDEV-444988 - Fix hiddenheap variable initialization
Change-Id: Icdb32ef4718b8ca0beedd569a4aaf40a9f4fb49d
2024-02-28 17:22:32 -05:00
Karl W Schulz 06f3fefa8e remove coverage reporting for now
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: 3111b7ef99]
2024-02-28 16:14:00 -06:00
Karl W Schulz 3111b7ef99 remove coverage reporting for now
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
2024-02-28 16:14:00 -06:00
German Andryeyev e26e82d58e SWDEV-311271 - Update mempool errors
Add extra validations for negative tests.

Change-Id: I7b2879b8840b5d1b6c5cb0d919a651cd868b7529


[ROCm/clr commit: d808ae6782]
2024-02-28 17:13:49 -05:00
German Andryeyev d808ae6782 SWDEV-311271 - Update mempool errors
Add extra validations for negative tests.

Change-Id: I7b2879b8840b5d1b6c5cb0d919a651cd868b7529
2024-02-28 17:13:49 -05:00
Karl W Schulz 0d6efcbc54 tweaking yaml syntax
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: 8c53982687]
2024-02-28 16:12:39 -06:00
Karl W Schulz 8c53982687 tweaking yaml syntax
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
2024-02-28 16:12:39 -06:00
Karl W Schulz 7168c9190d trying another coverage tool
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: 91c268e064]
2024-02-28 16:09:47 -06:00
Karl W Schulz 91c268e064 trying another coverage tool
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
2024-02-28 16:09:47 -06:00
Karl W Schulz 5f28f394da restore all profiling test definitions
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: f12794fadb]
2024-02-28 16:03:48 -06:00
Karl W Schulz f12794fadb restore all profiling test definitions
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
2024-02-28 16:03:48 -06:00
Karl W Schulz ca66997b32 enable yet another reporter
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: 91ea609670]
2024-02-28 15:55:49 -06:00
Karl W Schulz 91ea609670 enable yet another reporter
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
2024-02-28 15:55:49 -06:00