AI/rocm-systems
2
0 Форкнуть 0
Код Задачи Запросы на слияние Действия Пакеты Проекты Релизы Вики Активность
76 333 коммитов 3 Ветки 0 Теги
develop
Граф коммитов
76333 Коммитов
This Branch
This Branch
Все ветки
Автор SHA1 Сообщение Дата
foreman ddecac46a9 P4 to Git Change 1547064 by johtaylo@johtaylo-jtincrementor2-increment on 2018/04/27 03:00:07 
SWDEV-2 - Change OpenCL version number from 2640 to 2641.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2388 edit
 2018-04-27 03:07:27 -04:00
foreman 3e13e0b475 P4 to Git Change 1546955 by vsytchen@vsytchen-ocl-win10 on 2018/04/26 18:13:03 
SWDEV-151646 - [Win7_64][VEGA 10] Hard hang/blank screen observed while executing PC MARK8 (Video group chart)

	1. Revert to using max number of command buffers for device queue.

	ReviewBoardURL = http://ocltc.amd.com/reviews/r/14714/diff/

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#95 edit


[ROCm/clr commit: fa3460a862]
 2018-04-26 18:18:28 -04:00
foreman fa3460a862 P4 to Git Change 1546955 by vsytchen@vsytchen-ocl-win10 on 2018/04/26 18:13:03 
SWDEV-151646 - [Win7_64][VEGA 10] Hard hang/blank screen observed while executing PC MARK8 (Video group chart)

	1. Revert to using max number of command buffers for device queue.

	ReviewBoardURL = http://ocltc.amd.com/reviews/r/14714/diff/

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#95 edit
 2018-04-26 18:18:28 -04:00
foreman 32f7708794 P4 to Git Change 1546804 by skudchad@skudchad_test2_win_opencl on 2018/04/26 14:28:38 
SWDEV-145570 - [HIP] - Enable SVM mode if BUILD_HIP is defined.

	ReviewBoardURL = http://ocltc.amd.com/reviews/r/14711/diff/

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#87 edit


[ROCm/clr commit: ca538dfa45]
 2018-04-26 14:51:10 -04:00
foreman ca538dfa45 P4 to Git Change 1546804 by skudchad@skudchad_test2_win_opencl on 2018/04/26 14:28:38 
SWDEV-145570 - [HIP] - Enable SVM mode if BUILD_HIP is defined.

	ReviewBoardURL = http://ocltc.amd.com/reviews/r/14711/diff/

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#87 edit
 2018-04-26 14:51:10 -04:00
foreman dd4d5dcb94 P4 to Git Change 1546657 by gandryey@gera-w8 on 2018/04/26 10:59:34 
SWDEV-151739 - [CQE OCL][DTB][Perf][QR][DTB-BLOCKER][VEGA10] Upto 18% performance drop observed while running Video Composition test sub test of Compubench due to faulty CL#1544622
	- Implement customized TS tracking for managed buffers. The common TS tracking mechanism saves the event of the last command, assuming SDMA and compute operations occur in order, but for managed buffers it's not the case. Also managed buffer doesn't have to validate TS for the parent resource.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palblit.cpp#21 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palconstbuf.cpp#11 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palconstbuf.hpp#9 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palmemory.hpp#6 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palresource.hpp#22 edit


[ROCm/clr commit: 79ba5904dc]
 2018-04-26 11:13:29 -04:00
foreman 79ba5904dc P4 to Git Change 1546657 by gandryey@gera-w8 on 2018/04/26 10:59:34 
SWDEV-151739 - [CQE OCL][DTB][Perf][QR][DTB-BLOCKER][VEGA10] Upto 18% performance drop observed while running Video Composition test sub test of Compubench due to faulty CL#1544622
	- Implement customized TS tracking for managed buffers. The common TS tracking mechanism saves the event of the last command, assuming SDMA and compute operations occur in order, but for managed buffers it's not the case. Also managed buffer doesn't have to validate TS for the parent resource.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palblit.cpp#21 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palconstbuf.cpp#11 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palconstbuf.hpp#9 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palmemory.hpp#6 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palresource.hpp#22 edit
 2018-04-26 11:13:29 -04:00
foreman 7412bc74b3 P4 to Git Change 1546499 by johtaylo@johtaylo-jtincrementor2-increment on 2018/04/26 03:00:09 
SWDEV-2 - Change OpenCL version number from 2639 to 2640.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2387 edit


[ROCm/clr commit: d6618edf9f]
 2018-04-26 03:11:27 -04:00
foreman d6618edf9f P4 to Git Change 1546499 by johtaylo@johtaylo-jtincrementor2-increment on 2018/04/26 03:00:09 
SWDEV-2 - Change OpenCL version number from 2639 to 2640.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2387 edit
 2018-04-26 03:11:27 -04:00
Sean Keely 7cd6e366ed Workaround SDMA poll packet preemption. 
Use async. signal handler to satisfy dependencies for SDMA blits.

Change-Id: Ifa8d3ee6810509f400a568ca2387ac6ab3ab7c36
 2018-04-26 02:00:33 -05:00
Sean Keely 950dd4d89c Workaround SDMA poll packet preemption. 
Use async. signal handler to satisfy dependencies for SDMA blits.

Change-Id: Ifa8d3ee6810509f400a568ca2387ac6ab3ab7c36


[ROCm/ROCR-Runtime commit: 7cd6e366ed]
 2018-04-26 02:00:33 -05:00
Evgeny 0d286e5efa the timestamp to ns conversion correction 
Change-Id: I4928519dcef5850a9e43670d0e916e52ebf8cb0b
 2018-04-25 16:57:18 -05:00
Evgeny 68cfb95f04 the timestamp to ns conversion correction 
Change-Id: I4928519dcef5850a9e43670d0e916e52ebf8cb0b


[ROCm/rocprofiler commit: 0d286e5efa]
 2018-04-25 16:57:18 -05:00
Deven Desai 9946827d1b Checkin to add unit tests for fixes in my previous commit 
This change adds unit tests for

1. __hisinf
2. __hisnan
2. abs(long long)
3. lgamma(double)


[ROCm/clr commit: 76e808304c]
 2018-04-25 17:53:21 +00:00
Deven Desai 76e808304c Checkin to add unit tests for fixes in my previous commit 
This change adds unit tests for

1. __hisinf
2. __hisnan
2. abs(long long)
3. lgamma(double)
 2018-04-25 17:53:21 +00:00
Deven Desai 2b3332ac72 Checkin to add unit tests for fixes in my previous commit 
This change adds unit tests for

1. __hisinf
2. __hisnan
2. abs(long long)
3. lgamma(double)


[ROCm/hip commit: 4be4cf644f]
 2018-04-25 17:53:21 +00:00
Deven Desai 4be4cf644f Checkin to add unit tests for fixes in my previous commit 
This change adds unit tests for

1. __hisinf
2. __hisnan
2. abs(long long)
3. lgamma(double)
 2018-04-25 17:53:21 +00:00
foreman 4db0c47650 P4 to Git Change 1546224 by skudchad@skudchad_test2_win_opencl on 2018/04/25 12:24:40 
SWDEV-145570 - [HIP] - Add hipMemcpyTo/FromArray APIs.

	ReviewBoardURL = http://ocltc.amd.com/reviews/r/14696/diff/

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/hip/hip_memory.cpp#12 edit


[ROCm/clr commit: 9aef0311c1]
 2018-04-25 12:32:11 -04:00
foreman 9aef0311c1 P4 to Git Change 1546224 by skudchad@skudchad_test2_win_opencl on 2018/04/25 12:24:40 
SWDEV-145570 - [HIP] - Add hipMemcpyTo/FromArray APIs.

	ReviewBoardURL = http://ocltc.amd.com/reviews/r/14696/diff/

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/hip/hip_memory.cpp#12 edit
 2018-04-25 12:32:11 -04:00
foreman 67731a083a P4 to Git Change 1546224 by skudchad@skudchad_test2_win_opencl on 2018/04/25 12:24:40 
SWDEV-145570 - [HIP] - Add hipMemcpyTo/FromArray APIs.

	ReviewBoardURL = http://ocltc.amd.com/reviews/r/14696/diff/

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/hip/hip_memory.cpp#12 edit


[ROCm/hip commit: 18c7c164c6]
 2018-04-25 12:32:11 -04:00
foreman 18c7c164c6 P4 to Git Change 1546224 by skudchad@skudchad_test2_win_opencl on 2018/04/25 12:24:40 
SWDEV-145570 - [HIP] - Add hipMemcpyTo/FromArray APIs.

	ReviewBoardURL = http://ocltc.amd.com/reviews/r/14696/diff/

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/hip/hip_memory.cpp#12 edit
 2018-04-25 12:32:11 -04:00
foreman e4110c46b0 P4 to Git Change 1546206 by gandryey@gera-w8 on 2018/04/25 11:32:37 
SWDEV-151852 - [CQE OCL][DTB-BLOCKER][VEGA10][WIN][LNX][QR] Observed crash while running Multiple device context test due to FAULTY CL#1544622
	- Add a check for nullptr in case if virtual device object was destroyed before memory allocation

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palconstbuf.cpp#10 edit


[ROCm/clr commit: 3416ad385d]
 2018-04-25 11:40:42 -04:00
foreman 3416ad385d P4 to Git Change 1546206 by gandryey@gera-w8 on 2018/04/25 11:32:37 
SWDEV-151852 - [CQE OCL][DTB-BLOCKER][VEGA10][WIN][LNX][QR] Observed crash while running Multiple device context test due to FAULTY CL#1544622
	- Add a check for nullptr in case if virtual device object was destroyed before memory allocation

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palconstbuf.cpp#10 edit
 2018-04-25 11:40:42 -04:00
foreman 3e004a6860 P4 to Git Change 1546034 by johtaylo@johtaylo-jtincrementor2-increment on 2018/04/25 03:00:07 
SWDEV-2 - Change OpenCL version number from 2638 to 2639.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2386 edit


[ROCm/clr commit: b785abe728]
 2018-04-25 03:08:38 -04:00
foreman b785abe728 P4 to Git Change 1546034 by johtaylo@johtaylo-jtincrementor2-increment on 2018/04/25 03:00:07 
SWDEV-2 - Change OpenCL version number from 2638 to 2639.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2386 edit
 2018-04-25 03:08:38 -04:00
Rahul Garg f932a51c8e Fix gcc build issues for host code 
[ROCm/clr commit: 720e209366]
 2018-04-25 12:30:11 +05:30
Rahul Garg 720e209366 Fix gcc build issues for host code 2018-04-25 12:30:11 +05:30
Rahul Garg c1b94eb29f Fix gcc build issues for host code 
[ROCm/hip commit: 639001933c]
 2018-04-25 12:30:11 +05:30
Rahul Garg 639001933c Fix gcc build issues for host code 2018-04-25 12:30:11 +05:30
Evgeny 729ce983cb timestamp fix, converting to ns from system timestamp 2018-04-24 19:10:46 -05:00
Evgeny 71337f98ff timestamp fix, converting to ns from system timestamp 
[ROCm/rocprofiler commit: 729ce983cb]
 2018-04-24 19:10:46 -05:00
foreman 6a915378ee P4 to Git Change 1545906 by gandryey@gera-lnx-rcf on 2018/04/24 19:10:41 
SWDEV-79445 - OCL generic changes and code clean-up
	- Report an error if persistent memory allocation failed. The original logic was switching to local memory, but SSG can't work with it. The logic should work properly under Linux, but in Windows OS never fails persistent allocation.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#86 edit


[ROCm/clr commit: 1c91644f1b]
 2018-04-24 19:17:55 -04:00
foreman 1c91644f1b P4 to Git Change 1545906 by gandryey@gera-lnx-rcf on 2018/04/24 19:10:41 
SWDEV-79445 - OCL generic changes and code clean-up
	- Report an error if persistent memory allocation failed. The original logic was switching to local memory, but SSG can't work with it. The logic should work properly under Linux, but in Windows OS never fails persistent allocation.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#86 edit
 2018-04-24 19:17:55 -04:00
foreman dacb39cafa P4 to Git Change 1545895 by gandryey@gera-w8 on 2018/04/24 18:25:56 
SWDEV-79445 - OCL generic changes and code clean-up
	- Replace dynamic memory object allocation for the managed buffer with a preallocated object

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/device.hpp#298 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palconstbuf.cpp#9 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palconstbuf.hpp#8 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palmemory.hpp#5 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palresource.cpp#64 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palresource.hpp#21 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#94 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.hpp#50 edit


[ROCm/clr commit: 2f41f41977]
 2018-04-24 18:35:32 -04:00
foreman 2f41f41977 P4 to Git Change 1545895 by gandryey@gera-w8 on 2018/04/24 18:25:56 
SWDEV-79445 - OCL generic changes and code clean-up
	- Replace dynamic memory object allocation for the managed buffer with a preallocated object

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/device.hpp#298 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palconstbuf.cpp#9 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palconstbuf.hpp#8 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palmemory.hpp#5 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palresource.cpp#64 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palresource.hpp#21 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#94 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.hpp#50 edit
 2018-04-24 18:35:32 -04:00
foreman 92d104bfba P4 to Git Change 1545859 by cpaquot@cpaquot-ocl-lc-lnx on 2018/04/24 16:44:17 
SWDEV-145570 - [HIP] Get hipCtx_simple to pass
	Implemented hipCtxGetDevice
	hipCtxCreate must push the created context onto the context stack
	hipCtxDestroy must check if the top of the stack is the context being destroy
	and not just pop the top of the stack w/o checking.

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/hip/hip_context.cpp#8 edit


[ROCm/clr commit: 894f01c206]
 2018-04-24 16:49:38 -04:00
foreman 894f01c206 P4 to Git Change 1545859 by cpaquot@cpaquot-ocl-lc-lnx on 2018/04/24 16:44:17 
SWDEV-145570 - [HIP] Get hipCtx_simple to pass
	Implemented hipCtxGetDevice
	hipCtxCreate must push the created context onto the context stack
	hipCtxDestroy must check if the top of the stack is the context being destroy
	and not just pop the top of the stack w/o checking.

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/hip/hip_context.cpp#8 edit
 2018-04-24 16:49:38 -04:00
foreman 066a072226 P4 to Git Change 1545859 by cpaquot@cpaquot-ocl-lc-lnx on 2018/04/24 16:44:17 
SWDEV-145570 - [HIP] Get hipCtx_simple to pass
	Implemented hipCtxGetDevice
	hipCtxCreate must push the created context onto the context stack
	hipCtxDestroy must check if the top of the stack is the context being destroy
	and not just pop the top of the stack w/o checking.

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/hip/hip_context.cpp#8 edit


[ROCm/hip commit: d2fbde728c]
 2018-04-24 16:49:38 -04:00
foreman d2fbde728c P4 to Git Change 1545859 by cpaquot@cpaquot-ocl-lc-lnx on 2018/04/24 16:44:17 
SWDEV-145570 - [HIP] Get hipCtx_simple to pass
	Implemented hipCtxGetDevice
	hipCtxCreate must push the created context onto the context stack
	hipCtxDestroy must check if the top of the stack is the context being destroy
	and not just pop the top of the stack w/o checking.

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/hip/hip_context.cpp#8 edit
 2018-04-24 16:49:38 -04:00
Paul e4b8bec66d Add host and device targets 
[ROCm/clr commit: 56734306ed]
 2018-04-24 14:04:24 -05:00
Paul 56734306ed Add host and device targets 2018-04-24 14:04:24 -05:00
Paul 8643ed5ca9 Add host and device targets 
[ROCm/hip commit: 516d92f4d7]
 2018-04-24 14:04:24 -05:00
Paul 516d92f4d7 Add host and device targets 2018-04-24 14:04:24 -05:00
Evgeny Mankov 8426a907cb [HIPIFY] Sync with HIP (Execution Control, Surfaces, Memory) 
Execution Control [deprecated since 7.0]:
cudaConfigureCall -> hipConfigureCall
cudaLaunch -> hipLaunchByPtr
cudaSetupArgument -> hipSetupArgument

Surfaces (partially):
cudaSurfaceObject_t -> hipSurfaceObject_t
cudaSurfaceBoundaryMode -> hipSurfaceBoundaryMode
cudaCreateSurfaceObject -> hipCreateSurfaceObject
cudaDestroySurfaceObject -> hipDestroySurfaceObject

Memory Management:
cudaMemset2DAsync -> hipMemset2DAsync


[ROCm/clr commit: bfee03f95b]
 2018-04-24 22:04:04 +03:00
Evgeny Mankov bfee03f95b [HIPIFY] Sync with HIP (Execution Control, Surfaces, Memory) 
Execution Control [deprecated since 7.0]:
cudaConfigureCall -> hipConfigureCall
cudaLaunch -> hipLaunchByPtr
cudaSetupArgument -> hipSetupArgument

Surfaces (partially):
cudaSurfaceObject_t -> hipSurfaceObject_t
cudaSurfaceBoundaryMode -> hipSurfaceBoundaryMode
cudaCreateSurfaceObject -> hipCreateSurfaceObject
cudaDestroySurfaceObject -> hipDestroySurfaceObject

Memory Management:
cudaMemset2DAsync -> hipMemset2DAsync
 2018-04-24 22:04:04 +03:00
Evgeny Mankov 6b53512635 [HIPIFY] Sync with HIP (Execution Control, Surfaces, Memory) 
Execution Control [deprecated since 7.0]:
cudaConfigureCall -> hipConfigureCall
cudaLaunch -> hipLaunchByPtr
cudaSetupArgument -> hipSetupArgument

Surfaces (partially):
cudaSurfaceObject_t -> hipSurfaceObject_t
cudaSurfaceBoundaryMode -> hipSurfaceBoundaryMode
cudaCreateSurfaceObject -> hipCreateSurfaceObject
cudaDestroySurfaceObject -> hipDestroySurfaceObject

Memory Management:
cudaMemset2DAsync -> hipMemset2DAsync


[ROCm/hip commit: 6f501f6094]
 2018-04-24 22:04:04 +03:00
Evgeny Mankov 6f501f6094 [HIPIFY] Sync with HIP (Execution Control, Surfaces, Memory) 
Execution Control [deprecated since 7.0]:
cudaConfigureCall -> hipConfigureCall
cudaLaunch -> hipLaunchByPtr
cudaSetupArgument -> hipSetupArgument

Surfaces (partially):
cudaSurfaceObject_t -> hipSurfaceObject_t
cudaSurfaceBoundaryMode -> hipSurfaceBoundaryMode
cudaCreateSurfaceObject -> hipCreateSurfaceObject
cudaDestroySurfaceObject -> hipDestroySurfaceObject

Memory Management:
cudaMemset2DAsync -> hipMemset2DAsync
 2018-04-24 22:04:04 +03:00
Deven Desai 13ab5368d0 Checkin to fix bugs in math functions. 
This change fixes the following bugs that were discovered while debuggnig TF unit test failures (cwise_ops_test)

1. __hisinf and __hisnan routines
   Both had incorrect implementations.

2. abs
   A "long long" (64bit int) version was missing, resulting in the 32bit version being used for 64bit ints (which resulted in incorrect results, when the value passed in was outside the 32bit int range)

3. lgamma
  We seemed to have a custom version for the 'double' datatype (which was giving incorrect results). Replaced it with a call to the 'double' version of the underlying 'hc::precision_math::lgamma'


[ROCm/clr commit: 65a90c55e7]
 2018-04-24 18:10:07 +00:00
Deven Desai 65a90c55e7 Checkin to fix bugs in math functions. 
This change fixes the following bugs that were discovered while debuggnig TF unit test failures (cwise_ops_test)

1. __hisinf and __hisnan routines
   Both had incorrect implementations.

2. abs
   A "long long" (64bit int) version was missing, resulting in the 32bit version being used for 64bit ints (which resulted in incorrect results, when the value passed in was outside the 32bit int range)

3. lgamma
  We seemed to have a custom version for the 'double' datatype (which was giving incorrect results). Replaced it with a call to the 'double' version of the underlying 'hc::precision_math::lgamma'
 2018-04-24 18:10:07 +00:00
Deven Desai 16d2b986aa Checkin to fix bugs in math functions. 
This change fixes the following bugs that were discovered while debuggnig TF unit test failures (cwise_ops_test)

1. __hisinf and __hisnan routines
   Both had incorrect implementations.

2. abs
   A "long long" (64bit int) version was missing, resulting in the 32bit version being used for 64bit ints (which resulted in incorrect results, when the value passed in was outside the 32bit int range)

3. lgamma
  We seemed to have a custom version for the 'double' datatype (which was giving incorrect results). Replaced it with a call to the 'double' version of the underlying 'hc::precision_math::lgamma'


[ROCm/hip commit: af586bbbf2]
 2018-04-24 18:10:07 +00:00