76333 Коммитов
Автор SHA1 Сообщение Дата
Tao Sang ce1baf5d04 Support numa policy set by user
Add CL_MEM_FOLLOW_USER_NUMA_POLICY

Change-Id: I90a19dac7641827dff2ceb9ef8ae5f3467ed87a1


[ROCm/clr commit: 53264a8a4a]
2020-06-19 18:16:47 -04:00
Tao Sang 53264a8a4a Support numa policy set by user
Add CL_MEM_FOLLOW_USER_NUMA_POLICY

Change-Id: I90a19dac7641827dff2ceb9ef8ae5f3467ed87a1
2020-06-19 18:16:47 -04:00
Chris Freehill 4c94842508 Ensure no device mutexes are left held on shut_down
Also, fix TestMutualExclusion and TestEvtNofifReadWrite.
Previously, some of the normal SetUp function was not
being done for this test. In some cases, no DRM
devices are being found on the test machine. Skip
those.

Change-Id: Iaa5a257841eb459aa57491ae9680c34a60d5ac2b


[ROCm/rocm_smi_lib commit: 59394f3354]
2020-06-19 13:59:20 -05:00
Chris Freehill 59394f3354 Ensure no device mutexes are left held on shut_down
Also, fix TestMutualExclusion and TestEvtNofifReadWrite.
Previously, some of the normal SetUp function was not
being done for this test. In some cases, no DRM
devices are being found on the test machine. Skip
those.

Change-Id: Iaa5a257841eb459aa57491ae9680c34a60d5ac2b
2020-06-19 13:59:20 -05:00
Chris Freehill 51d5c02343 Ensure no device mutexes are left held on shut_down
Also, fix TestMutualExclusion and TestEvtNofifReadWrite.
Previously, some of the normal SetUp function was not
being done for this test. In some cases, no DRM
devices are being found on the test machine. Skip
those.

Change-Id: Iaa5a257841eb459aa57491ae9680c34a60d5ac2b


[ROCm/amdsmi commit: 59394f3354]
2020-06-19 13:59:20 -05:00
Evgeny 6c72cdfe82 json trace normalizing fix
Change-Id: Ia5987fcd3e08808953534fa3ded2df0b17e0e87a
2020-06-19 13:40:33 -05:00
Evgeny 5649332eb5 json trace normalizing fix
Change-Id: Ia5987fcd3e08808953534fa3ded2df0b17e0e87a


[ROCm/rocprofiler commit: 6c72cdfe82]
2020-06-19 13:40:33 -05:00
Evgeny 4e299dbfa4 fixing ostreams linking as inline static
Change-Id: Idbfac3e5220acd62500cd4c204a5bd9a9461c8fc


[ROCm/roctracer commit: 264e3d816d]
2020-06-19 12:54:12 -05:00
Evgeny 264e3d816d fixing ostreams linking as inline static
Change-Id: Idbfac3e5220acd62500cd4c204a5bd9a9461c8fc
2020-06-19 12:54:12 -05:00
Sylvain Jeaugey 12d86bd58f Change all_gather/reduce_scatter algbw to match the documentation.
Fix #45 : All_gather and reduce_scatter algorithm bandwidth was
computed as time/count*(nranks-1) which is not consistent with the
way we compute it for other collectives.

This change makes algbw higher; busbw is unchanged.


[ROCm/rccl-tests commit: ec1b5e22e6]
2020-06-19 10:42:19 -07:00
Sylvain Jeaugey ec1b5e22e6 Change all_gather/reduce_scatter algbw to match the documentation.
Fix #45 : All_gather and reduce_scatter algorithm bandwidth was
computed as time/count*(nranks-1) which is not consistent with the
way we compute it for other collectives.

This change makes algbw higher; busbw is unchanged.
2020-06-19 10:42:19 -07:00
Jason Tang f1110d715e Set the tiled_ flag according to ImageCreateInfo
Change-Id: I3e1b2fdd15e32c3c24f4341001b068738c98e388


[ROCm/clr commit: 138b2d3ce2]
2020-06-19 12:34:07 -04:00
Jason Tang 138b2d3ce2 Set the tiled_ flag according to ImageCreateInfo
Change-Id: I3e1b2fdd15e32c3c24f4341001b068738c98e388
2020-06-19 12:34:07 -04:00
kjayapra-amd 5a1e09a37d SWDEV-229840 - Add Debug prints when the element is already present in MemObjMap_
Change-Id: I21129d087e73cc2a9e35f03e6a1a2dc527626f48


[ROCm/clr commit: 71c05075ba]
2020-06-19 12:33:20 -04:00
kjayapra-amd 71c05075ba SWDEV-229840 - Add Debug prints when the element is already present in MemObjMap_
Change-Id: I21129d087e73cc2a9e35f03e6a1a2dc527626f48
2020-06-19 12:33:20 -04:00
German Andryeyev a1974c06e9 Switch PAL to interface 610
Change-Id: Ida563f38513789f5ba105ac92a9737c59cd5d90d


[ROCm/clr commit: 01fcf82448]
2020-06-19 11:40:12 -04:00
German Andryeyev 01fcf82448 Switch PAL to interface 610
Change-Id: Ida563f38513789f5ba105ac92a9737c59cd5d90d
2020-06-19 11:40:12 -04:00
Vlad Sytchenko 02641f4f43 Correct texture element size calculations
hipChannelFormatDesc contains size of each channel in bits. There's 8 bits per byte, not 4, so we should divide by 8.

Change-Id: I38ae2d28ad54740823c98908ca2afecad0ba0183


[ROCm/clr commit: 81b15ed36e]
2020-06-19 11:39:03 -04:00
Vlad Sytchenko 81b15ed36e Correct texture element size calculations
hipChannelFormatDesc contains size of each channel in bits. There's 8 bits per byte, not 4, so we should divide by 8.

Change-Id: I38ae2d28ad54740823c98908ca2afecad0ba0183
2020-06-19 11:39:03 -04:00
Vlad Sytchenko 2780b86df2 Correct texture element size calculations
hipChannelFormatDesc contains size of each channel in bits. There's 8 bits per byte, not 4, so we should divide by 8.

Change-Id: I38ae2d28ad54740823c98908ca2afecad0ba0183


[ROCm/hip commit: e9033edc10]
2020-06-19 11:39:03 -04:00
Vlad Sytchenko e9033edc10 Correct texture element size calculations
hipChannelFormatDesc contains size of each channel in bits. There's 8 bits per byte, not 4, so we should divide by 8.

Change-Id: I38ae2d28ad54740823c98908ca2afecad0ba0183
2020-06-19 11:39:03 -04:00
Jason Tang fb0be5d639 Merge pal Resource mapLayers/unmapLayers logic into map/unmap
Change-Id: I0198fb96f104539fb557f91160665e60ba576132


[ROCm/clr commit: 155ff5bdcc]
2020-06-19 10:51:48 -04:00
Jason Tang 155ff5bdcc Merge pal Resource mapLayers/unmapLayers logic into map/unmap
Change-Id: I0198fb96f104539fb557f91160665e60ba576132
2020-06-19 10:51:48 -04:00
Mike Li b3bb190b8d Add support to retrieve XGMI hive id
Change-Id: I1eee05dd85ecb856889d1cfe0565454d2f538856
Signed-off-by: Mike Li <Tianxinmike.Li@amd.com>


[ROCm/rocm_smi_lib commit: 488bbb668a]
2020-06-19 07:35:23 -07:00
Mike Li 488bbb668a Add support to retrieve XGMI hive id
Change-Id: I1eee05dd85ecb856889d1cfe0565454d2f538856
Signed-off-by: Mike Li <Tianxinmike.Li@amd.com>
2020-06-19 07:35:23 -07:00
Mike Li 9375ef92fc Add support to retrieve XGMI hive id
Change-Id: I1eee05dd85ecb856889d1cfe0565454d2f538856
Signed-off-by: Mike Li <Tianxinmike.Li@amd.com>


[ROCm/amdsmi commit: 488bbb668a]
2020-06-19 07:35:23 -07:00
Sylvain Jeaugey fcaaf2c4a1 Fix #47 : compilation error on NCCL<2.7
Return an error when trying to run alltoall test when compiled
against NCCL<2.7.


[ROCm/rccl-tests commit: 07ac716c1a]
2020-06-18 15:02:51 -07:00
Sylvain Jeaugey 07ac716c1a Fix #47 : compilation error on NCCL<2.7
Return an error when trying to run alltoall test when compiled
against NCCL<2.7.
2020-06-18 15:02:51 -07:00
Yaxun (Sam) Liu 8849a591f7 Fix missing ldexp(float,int)
Change-Id: I2c1553407dfc26948d3ab7aa532eef42a0f6b204


[ROCm/clr commit: 2461135d8a]
2020-06-18 15:16:59 -04:00
Yaxun (Sam) Liu 2461135d8a Fix missing ldexp(float,int)
Change-Id: I2c1553407dfc26948d3ab7aa532eef42a0f6b204
2020-06-18 15:16:59 -04:00
Yaxun (Sam) Liu 7f26819b2f Fix missing ldexp(float,int)
Change-Id: I2c1553407dfc26948d3ab7aa532eef42a0f6b204


[ROCm/hip commit: b907505d55]
2020-06-18 15:16:59 -04:00
Yaxun (Sam) Liu b907505d55 Fix missing ldexp(float,int)
Change-Id: I2c1553407dfc26948d3ab7aa532eef42a0f6b204
2020-06-18 15:16:59 -04:00
Aryan Salmanpour 6943b4fa83 move hostcall-based implementation of messages/printf from rocm backend to the common device layer
SWDEV-232428

Change-Id: I14b774e66a1783569c6d34d8759c1faa5f340e27


[ROCm/clr commit: 1bd265f8d6]
2020-06-18 13:49:28 -04:00
Aryan Salmanpour 1bd265f8d6 move hostcall-based implementation of messages/printf from rocm backend to the common device layer
SWDEV-232428

Change-Id: I14b774e66a1783569c6d34d8759c1faa5f340e27
2020-06-18 13:49:28 -04:00
rohit pathania 11ef0984f4 [ dtest ] Added different test scenario with hipMemcpyWithStream
- Test with one, two streams
- Test Multi-gpu (one stream per gpu)
- Test D-D (on same device/different devices).  Can discover issues
  when devices are on same or different root complexes.
- Test H-D/Default

SWDEV-238517 for enhancing hip unit tests

Change-Id: I8031a7eebe2f9c8c0e0996e2c7accb09ac0b96d4


[ROCm/clr commit: 226c438559]
2020-06-18 10:31:01 -04:00
rohit pathania 226c438559 [ dtest ] Added different test scenario with hipMemcpyWithStream
- Test with one, two streams
- Test Multi-gpu (one stream per gpu)
- Test D-D (on same device/different devices).  Can discover issues
  when devices are on same or different root complexes.
- Test H-D/Default

SWDEV-238517 for enhancing hip unit tests

Change-Id: I8031a7eebe2f9c8c0e0996e2c7accb09ac0b96d4
2020-06-18 10:31:01 -04:00
rohit pathania 670a34e8e2 [ dtest ] Added different test scenario with hipMemcpyWithStream
- Test with one, two streams
- Test Multi-gpu (one stream per gpu)
- Test D-D (on same device/different devices).  Can discover issues
  when devices are on same or different root complexes.
- Test H-D/Default

SWDEV-238517 for enhancing hip unit tests

Change-Id: I8031a7eebe2f9c8c0e0996e2c7accb09ac0b96d4


[ROCm/hip commit: d701632e01]
2020-06-18 10:31:01 -04:00
rohit pathania d701632e01 [ dtest ] Added different test scenario with hipMemcpyWithStream
- Test with one, two streams
- Test Multi-gpu (one stream per gpu)
- Test D-D (on same device/different devices).  Can discover issues
  when devices are on same or different root complexes.
- Test H-D/Default

SWDEV-238517 for enhancing hip unit tests

Change-Id: I8031a7eebe2f9c8c0e0996e2c7accb09ac0b96d4
2020-06-18 10:31:01 -04:00
Chauncey Hui 8153018c23 SWDEV-2 - Change OpenCL version number from 3164 to 3165
[ROCm/clr commit: c9c4250bbe]
2020-06-18 03:00:03 -04:00
Chauncey Hui c9c4250bbe SWDEV-2 - Change OpenCL version number from 3164 to 3165 2020-06-18 03:00:03 -04:00
Laurent Morichetti db6a781f0c Preserve the exec mask for the queue error case
Save and restore exec_lo/exec_hi around the mGetDoorbellId macro in
the signal_error case just like we do in the signal_debugger case.

Also reset the wave_id (ttmp4/ttmp5) to 0 since it isn't preserved.
0 will be detected as a new wave by the debugger api library.

Change-Id: I5123caa9431154ec1584bae85e42648c97c64c37
2020-06-18 02:25:11 -04:00
Laurent Morichetti f80ce871b3 Preserve the exec mask for the queue error case
Save and restore exec_lo/exec_hi around the mGetDoorbellId macro in
the signal_error case just like we do in the signal_debugger case.

Also reset the wave_id (ttmp4/ttmp5) to 0 since it isn't preserved.
0 will be detected as a new wave by the debugger api library.

Change-Id: I5123caa9431154ec1584bae85e42648c97c64c37


[ROCm/ROCR-Runtime commit: db6a781f0c]
2020-06-18 02:25:11 -04:00
Chris Freehill 4fd855fe72 Fix line endings for init_shutdown_refcount.*
* Also, add assert that check for proper usage of
rand_sleep_mod().

Change-Id: Ieb4179e1ad12fbbf85c2e4f7c7f119b0bb30b197


[ROCm/rocm_smi_lib commit: 9e0ebb250c]
2020-06-17 21:26:12 -05:00
Chris Freehill 9e0ebb250c Fix line endings for init_shutdown_refcount.*
* Also, add assert that check for proper usage of
rand_sleep_mod().

Change-Id: Ieb4179e1ad12fbbf85c2e4f7c7f119b0bb30b197
2020-06-17 21:26:12 -05:00
Chris Freehill 1309615850 Fix line endings for init_shutdown_refcount.*
* Also, add assert that check for proper usage of
rand_sleep_mod().

Change-Id: Ieb4179e1ad12fbbf85c2e4f7c7f119b0bb30b197


[ROCm/amdsmi commit: 9e0ebb250c]
2020-06-17 21:26:12 -05:00
Chris Freehill b1787b8968 Make verbosity level 0 completely quiet
Also, support --iterations flag for certain functions that will
likely be repeated frequently.

Change-Id: I7ed76835001b5cbca30042d6bf26484258c7b9a6


[ROCm/rocm_smi_lib commit: efc9b7658c]
2020-06-17 21:26:12 -05:00
Chris Freehill efc9b7658c Make verbosity level 0 completely quiet
Also, support --iterations flag for certain functions that will
likely be repeated frequently.

Change-Id: I7ed76835001b5cbca30042d6bf26484258c7b9a6
2020-06-17 21:26:12 -05:00
Chris Freehill 1739bf2777 Make verbosity level 0 completely quiet
Also, support --iterations flag for certain functions that will
likely be repeated frequently.

Change-Id: I7ed76835001b5cbca30042d6bf26484258c7b9a6


[ROCm/amdsmi commit: efc9b7658c]
2020-06-17 21:26:12 -05:00
Evgeny dfba287ff2 json trace normalizing for app start
Change-Id: I97480f7461c377096c101cd0dc6ce088f799fbab
2020-06-17 19:01:12 -05:00
Evgeny 80ad421ce0 json trace normalizing for app start
Change-Id: I97480f7461c377096c101cd0dc6ce088f799fbab


[ROCm/rocprofiler commit: dfba287ff2]
2020-06-17 19:01:12 -05:00