76333 Коммитов
Автор SHA1 Сообщение Дата
Evgeny Mankov 1c0b4b6c13 Merge pull request #854 from emankov/master
[HIPIFY][cmake] Add cmake versioning depending on CUDA version

[ROCm/hip commit: fb0877564e]
2019-01-09 16:39:30 +03:00
Evgeny Mankov fb0877564e Merge pull request #854 from emankov/master
[HIPIFY][cmake] Add cmake versioning depending on CUDA version
2019-01-09 16:39:30 +03:00
Evgeny Mankov 587969c10e Merge branch 'master' of https://github.com/ROCm-Developer-Tools/HIP
[ROCm/clr commit: 5dba4e74cb]
2019-01-09 14:03:40 +03:00
Evgeny Mankov 5dba4e74cb Merge branch 'master' of https://github.com/ROCm-Developer-Tools/HIP 2019-01-09 14:03:40 +03:00
Evgeny Mankov 643f84476c Merge branch 'master' of https://github.com/ROCm-Developer-Tools/HIP
[ROCm/hip commit: f5effc0d23]
2019-01-09 14:03:40 +03:00
Evgeny Mankov f5effc0d23 Merge branch 'master' of https://github.com/ROCm-Developer-Tools/HIP 2019-01-09 14:03:40 +03:00
Evgeny Mankov eb42f06a52 [HIPIFY][cmake] Add cmake versioning depending on CUDA version
[ROCm/clr commit: d160c7285c]
2019-01-09 13:59:48 +03:00
Evgeny Mankov d160c7285c [HIPIFY][cmake] Add cmake versioning depending on CUDA version 2019-01-09 13:59:48 +03:00
Evgeny Mankov e816f3e98a [HIPIFY][cmake] Add cmake versioning depending on CUDA version
[ROCm/hip commit: 8fb8ca02c8]
2019-01-09 13:59:48 +03:00
Evgeny Mankov 8fb8ca02c8 [HIPIFY][cmake] Add cmake versioning depending on CUDA version 2019-01-09 13:59:48 +03:00
Maneesh Gupta 4fdba912db Merge pull request #853 from ROCm-Developer-Tools/hip-module-hash-fix
Introduce hash key to HIP module implementation

[ROCm/clr commit: d3a5502f17]
2019-01-09 16:28:50 +05:30
Maneesh Gupta d3a5502f17 Merge pull request #853 from ROCm-Developer-Tools/hip-module-hash-fix
Introduce hash key to HIP module implementation
2019-01-09 16:28:50 +05:30
Maneesh Gupta ec245b61e9 Merge pull request #853 from ROCm-Developer-Tools/hip-module-hash-fix
Introduce hash key to HIP module implementation

[ROCm/hip commit: 36039d6765]
2019-01-09 16:28:50 +05:30
Maneesh Gupta 36039d6765 Merge pull request #853 from ROCm-Developer-Tools/hip-module-hash-fix
Introduce hash key to HIP module implementation
2019-01-09 16:28:50 +05:30
foreman eaa1f3b1e8 P4 to Git Change 1727825 by chui@ocl-promo-incrementor on 2019/01/09 03:00:06
SWDEV-2 - Change OpenCL version number from 2811 to 2812.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2559 edit


[ROCm/clr commit: ca14911693]
2019-01-09 03:12:22 -05:00
foreman ca14911693 P4 to Git Change 1727825 by chui@ocl-promo-incrementor on 2019/01/09 03:00:06
SWDEV-2 - Change OpenCL version number from 2811 to 2812.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2559 edit
2019-01-09 03:12:22 -05:00
Wen-Heng (Jack) Chung 2f64c76789 Introduce hash key to HIP module implementation
A hash calculated via FNV-1a algorithm is introduced in ihipModule_t, the
internal of hipModule_t. The hash is used by HIP module APIs such as

- read_agent_global_from_module

to determine whether the agent-scope globals for a module have been iterated.

This commit fixes one issue that applications which load / unload modules
frequently would occasionally fail. After deep investigation of the issue it
turns out the old implementation in read_agent_global_from_module uses
hipModule_t as the key, which is not robust enough, as hipModule_t instances
are allocated dynamically so there are cases that one memory address may be
used by multiple hipModule_t instances. The real solution is to introduce a
uniquely identifiable hash for the code object associated with the HIP module.
And that's the rationale behind this commit.


[ROCm/clr commit: 6e68d44220]
2019-01-08 17:33:40 +00:00
Wen-Heng (Jack) Chung 6e68d44220 Introduce hash key to HIP module implementation
A hash calculated via FNV-1a algorithm is introduced in ihipModule_t, the
internal of hipModule_t. The hash is used by HIP module APIs such as

- read_agent_global_from_module

to determine whether the agent-scope globals for a module have been iterated.

This commit fixes one issue that applications which load / unload modules
frequently would occasionally fail. After deep investigation of the issue it
turns out the old implementation in read_agent_global_from_module uses
hipModule_t as the key, which is not robust enough, as hipModule_t instances
are allocated dynamically so there are cases that one memory address may be
used by multiple hipModule_t instances. The real solution is to introduce a
uniquely identifiable hash for the code object associated with the HIP module.
And that's the rationale behind this commit.
2019-01-08 17:33:40 +00:00
Wen-Heng (Jack) Chung 2f9021afd3 Introduce hash key to HIP module implementation
A hash calculated via FNV-1a algorithm is introduced in ihipModule_t, the
internal of hipModule_t. The hash is used by HIP module APIs such as

- read_agent_global_from_module

to determine whether the agent-scope globals for a module have been iterated.

This commit fixes one issue that applications which load / unload modules
frequently would occasionally fail. After deep investigation of the issue it
turns out the old implementation in read_agent_global_from_module uses
hipModule_t as the key, which is not robust enough, as hipModule_t instances
are allocated dynamically so there are cases that one memory address may be
used by multiple hipModule_t instances. The real solution is to introduce a
uniquely identifiable hash for the code object associated with the HIP module.
And that's the rationale behind this commit.


[ROCm/hip commit: b4d658a48f]
2019-01-08 17:33:40 +00:00
Wen-Heng (Jack) Chung b4d658a48f Introduce hash key to HIP module implementation
A hash calculated via FNV-1a algorithm is introduced in ihipModule_t, the
internal of hipModule_t. The hash is used by HIP module APIs such as

- read_agent_global_from_module

to determine whether the agent-scope globals for a module have been iterated.

This commit fixes one issue that applications which load / unload modules
frequently would occasionally fail. After deep investigation of the issue it
turns out the old implementation in read_agent_global_from_module uses
hipModule_t as the key, which is not robust enough, as hipModule_t instances
are allocated dynamically so there are cases that one memory address may be
used by multiple hipModule_t instances. The real solution is to introduce a
uniquely identifiable hash for the code object associated with the HIP module.
And that's the rationale behind this commit.
2019-01-08 17:33:40 +00:00
foreman 09d6e70f9f P4 to Git Change 1727393 by gandryey@gera-w8 on 2019/01/08 11:35:08
SWDEV-79445 - OCL generic changes and code clean-up
	- Allow LC to execute 64bit GPU binaries with 32 bit OCL runtime.

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_program.cpp#51 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#122 edit


[ROCm/clr commit: 3574cc00d3]
2019-01-08 12:13:48 -05:00
foreman 3574cc00d3 P4 to Git Change 1727393 by gandryey@gera-w8 on 2019/01/08 11:35:08
SWDEV-79445 - OCL generic changes and code clean-up
	- Allow LC to execute 64bit GPU binaries with 32 bit OCL runtime.

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_program.cpp#51 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#122 edit
2019-01-08 12:13:48 -05:00
foreman 14e6355cd0 P4 to Git Change 1727393 by gandryey@gera-w8 on 2019/01/08 11:35:08
SWDEV-79445 - OCL generic changes and code clean-up
	- Allow LC to execute 64bit GPU binaries with 32 bit OCL runtime.

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_program.cpp#51 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#122 edit


[ROCm/clr commit: 697a5c68ca]
2019-01-08 12:13:48 -05:00
foreman 697a5c68ca P4 to Git Change 1727393 by gandryey@gera-w8 on 2019/01/08 11:35:08
SWDEV-79445 - OCL generic changes and code clean-up
	- Allow LC to execute 64bit GPU binaries with 32 bit OCL runtime.

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_program.cpp#51 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#122 edit
2019-01-08 12:13:48 -05:00
foreman 76a4a035fa P4 to Git Change 1727221 by chui@ocl-promo-incrementor on 2019/01/08 03:00:04
SWDEV-2 - Change OpenCL version number from 2810 to 2811.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2558 edit


[ROCm/clr commit: 0c6cf47532]
2019-01-08 03:14:11 -05:00
foreman 0c6cf47532 P4 to Git Change 1727221 by chui@ocl-promo-incrementor on 2019/01/08 03:00:04
SWDEV-2 - Change OpenCL version number from 2810 to 2811.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2558 edit
2019-01-08 03:14:11 -05:00
Evgeny 3e37e001fa minor change
[ROCm/roctracer commit: 5272bcc316]
2019-01-07 18:58:01 -06:00
Evgeny 5272bcc316 minor change 2019-01-07 18:58:01 -06:00
foreman 489260bb0d P4 to Git Change 1727059 by asalmanp@asalmanp-ocl-stg on 2019/01/07 17:48:44
SWDEV-132899 - [OCL][GFX10] report number of WGP by default on gfx10 ASICs
	Both HSAIL/SC and LC compilers use WGP mode by default on gfx10 ASICs (i.e., COMPUTE_PGM_RSRC1.WGP_MODE is set to 1 by both compilers) therefore runtime should report number of WGP (i.e., CU/2) on gfx10 ASICs by default.
	The new environment variable (GPU_ENABLE_WGP_MODE = 0)  can be used to force CU mode on LC (i.e., -mcumode option) if it’s needed (HSAIL/SC doesn't have any compiler option for forcing the CU mode)

	Also, using the new environment variable (GPU_ENABLE_WAVE32_MODE) to control the wave32 mode on gfx10+.

	ReviewRequestURL = http://ocltc.amd.com/reviews/r/16435/diff/

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/device.hpp#329 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/devprogram.cpp#27 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#121 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palsettings.cpp#65 edit
... //depot/stg/opencl/drivers/opencl/runtime/utils/flags.hpp#301 edit


[ROCm/clr commit: e631b3978f]
2019-01-07 18:02:10 -05:00
foreman e631b3978f P4 to Git Change 1727059 by asalmanp@asalmanp-ocl-stg on 2019/01/07 17:48:44
SWDEV-132899 - [OCL][GFX10] report number of WGP by default on gfx10 ASICs
	Both HSAIL/SC and LC compilers use WGP mode by default on gfx10 ASICs (i.e., COMPUTE_PGM_RSRC1.WGP_MODE is set to 1 by both compilers) therefore runtime should report number of WGP (i.e., CU/2) on gfx10 ASICs by default.
	The new environment variable (GPU_ENABLE_WGP_MODE = 0)  can be used to force CU mode on LC (i.e., -mcumode option) if it’s needed (HSAIL/SC doesn't have any compiler option for forcing the CU mode)

	Also, using the new environment variable (GPU_ENABLE_WAVE32_MODE) to control the wave32 mode on gfx10+.

	ReviewRequestURL = http://ocltc.amd.com/reviews/r/16435/diff/

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/device.hpp#329 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/devprogram.cpp#27 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#121 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palsettings.cpp#65 edit
... //depot/stg/opencl/drivers/opencl/runtime/utils/flags.hpp#301 edit
2019-01-07 18:02:10 -05:00
Christian Sigg 3a7974d4c4 Fix memory leak in bootstrapRoot()
[ROCm/rccl commit: 4861e197fd]
2019-01-07 14:18:46 -08:00
Christian Sigg 4861e197fd Fix memory leak in bootstrapRoot() 2019-01-07 14:18:46 -08:00
Chris Freehill 3234808d49 Add support for reading frequency-volt curve data
[ROCm/rocm_smi_lib commit: 64eec4673a]
2019-01-07 10:20:53 -06:00
Chris Freehill 64eec4673a Add support for reading frequency-volt curve data 2019-01-07 10:20:53 -06:00
Chris Freehill e3104f3ac0 Add support for reading frequency-volt curve data
[ROCm/amdsmi commit: 64eec4673a]
2019-01-07 10:20:53 -06:00
Chris Freehill 80eeb7960f Add support for reading frequency-volt curva data
[ROCm/rocm_smi_lib commit: 639a4e3503]
2019-01-07 08:44:23 -06:00
Chris Freehill 639a4e3503 Add support for reading frequency-volt curva data 2019-01-07 08:44:23 -06:00
Chris Freehill f2298dc9c1 Add support for reading frequency-volt curva data
[ROCm/amdsmi commit: 639a4e3503]
2019-01-07 08:44:23 -06:00
Evgeny 2b2a279ce9 hsa memcopy tracking
Change-Id: I61bc4ea01a62b587d5941f9d0e9e358a6ed71b5a
2019-01-07 01:53:27 -06:00
Evgeny 63150d03e1 hsa memcopy tracking
Change-Id: I61bc4ea01a62b587d5941f9d0e9e358a6ed71b5a


[ROCm/rocprofiler commit: 2b2a279ce9]
2019-01-07 01:53:27 -06:00
Maneesh Gupta d41c852742 Merge pull request #845 from mangupta/cleanup-install-markdown
Update INSTALL.md

[ROCm/clr commit: cf7dc26851]
2019-01-07 10:04:36 +05:30
Maneesh Gupta cf7dc26851 Merge pull request #845 from mangupta/cleanup-install-markdown
Update INSTALL.md
2019-01-07 10:04:36 +05:30
Maneesh Gupta 685b3b2167 Merge pull request #845 from mangupta/cleanup-install-markdown
Update INSTALL.md

[ROCm/hip commit: 7637504776]
2019-01-07 10:04:36 +05:30
Maneesh Gupta 7637504776 Merge pull request #845 from mangupta/cleanup-install-markdown
Update INSTALL.md
2019-01-07 10:04:36 +05:30
Maneesh Gupta 7b45721b1b Merge pull request #701 from ssahasra/stream-tests
introduce tests for asynchronous streams

[ROCm/clr commit: 03e1058666]
2019-01-07 10:04:29 +05:30
Maneesh Gupta 03e1058666 Merge pull request #701 from ssahasra/stream-tests
introduce tests for asynchronous streams
2019-01-07 10:04:29 +05:30
Maneesh Gupta 419f970c01 Merge pull request #701 from ssahasra/stream-tests
introduce tests for asynchronous streams

[ROCm/hip commit: 87d0674a1a]
2019-01-07 10:04:29 +05:30
Maneesh Gupta 87d0674a1a Merge pull request #701 from ssahasra/stream-tests
introduce tests for asynchronous streams
2019-01-07 10:04:29 +05:30
Maneesh Gupta e7195f7dc2 Update hipAsynchronousStreams.cpp
This test needs to be excluded from nvcc platform

[ROCm/clr commit: 79251fc4e2]
2019-01-07 09:26:21 +05:30
Maneesh Gupta 79251fc4e2 Update hipAsynchronousStreams.cpp
This test needs to be excluded from nvcc platform
2019-01-07 09:26:21 +05:30