Adds hsa_amd_portable_export_dmabuf and hsa_amd_portable_close_dmabuf
which allow obtaining dmabuf handles to rocr allocations. These handles
may be shared with other APIs to support cross vendor & cross device
memory sharing.
Adds query to return whether dmabuf export is supported
Signed-off-by: Jonathan Kim <Jonathan.Kim@amd.com>
Signed-off-by: David Yat Sin <David.YatSin@amd.com>
Change-Id: I7f98501087d9563d07fc2cb428cc886b1e518b1e
Adds hsa_amd_portable_export_dmabuf and hsa_amd_portable_close_dmabuf
which allow obtaining dmabuf handles to rocr allocations. These handles
may be shared with other APIs to support cross vendor & cross device
memory sharing.
Adds query to return whether dmabuf export is supported
Signed-off-by: Jonathan Kim <Jonathan.Kim@amd.com>
Signed-off-by: David Yat Sin <David.YatSin@amd.com>
Change-Id: I7f98501087d9563d07fc2cb428cc886b1e518b1e
[ROCm/ROCR-Runtime commit: 42243c1e8f]
The process ids of other container are still visible in the sysfs file,
filter it out to prevent crash.
Change-Id: I665912cd09c606804186aff8cba5c24f5e58ded7
[ROCm/rocm_smi_lib commit: 710649ab66]
The process ids of other container are still visible in the sysfs file,
filter it out to prevent crash.
Change-Id: I665912cd09c606804186aff8cba5c24f5e58ded7
* EXSWHTEC-171 - Implement positive and negative unit tests for the following API's:
- hipGraphCreate
- hipGraphDestroy
[ROCm/hip-tests commit: 42ebc6cd10]
- Implement positive and negative tests for hipGraphAddMemsetNode
- Implement positive and negative tests for hipGraphMemsetNodeSetParams
- Implement positive and negative tests for hipGraphExecMemsetNodeSetParams
- Implement negative tests for hipGraphMemsetNodeGetParams
- Add doxygen annotations
- Add a dynamic section in positive test for memset for different dimensions
[ROCm/hip-tests commit: 77a37371b0]
- Implement positive and negative tests for hipGraphAddMemsetNode
- Implement positive and negative tests for hipGraphMemsetNodeSetParams
- Implement positive and negative tests for hipGraphExecMemsetNodeSetParams
- Implement negative tests for hipGraphMemsetNodeGetParams
- Add doxygen annotations
- Add a dynamic section in positive test for memset for different dimensions
Implement tests for hipStreamUpdateCaptureDependencies, hipThreadExchangeStreamCaptureMode and hipLaunchHostFunc
- Refactor existing tests by including more catch2 features
- Reduce code line numbers by using helper guard classes
- Add some positive and negative tests
- Add doxygen test descriptions
[ROCm/hip-tests commit: 4580864fa1]
Implement tests for hipStreamUpdateCaptureDependencies, hipThreadExchangeStreamCaptureMode and hipLaunchHostFunc
- Refactor existing tests by including more catch2 features
- Reduce code line numbers by using helper guard classes
- Add some positive and negative tests
- Add doxygen test descriptions
- Refactor existing tests by including more catch2 features
- Reduce code line numbers by using helper guard classes
- Add some positive and negative tests
- Add doxygen test descriptions
[ROCm/hip-tests commit: 638e2aabeb]
- Refactor existing tests by including more catch2 features
- Reduce code line numbers by using helper guard classes
- Add some positive and negative tests
- Add doxygen test descriptions
* EXSWHTEC-192 - Implement new and update existing tests for the hipGraph*MemcpyNode[From|To]Symbol family of APIs
- Generalize from symbol positive test to work for different types
- Implement negative parameter tests
- Extract common code for to/from copy
- Reorganize files
[ROCm/hip-tests commit: f1151e3c6d]
* EXSWHTEC-192 - Implement new and update existing tests for the hipGraph*MemcpyNode[From|To]Symbol family of APIs
- Generalize from symbol positive test to work for different types
- Implement negative parameter tests
- Extract common code for to/from copy
- Reorganize files
Implemented positive and negative test cases for the following APIs:
- hipApiName
- hipKernelNameRef
- hipKernelNameRefByPtr
- hipGetStreamDeviceId
Enabled build of Callback tests only for AMD platforms, as there is no support for Nvidia platforms.
[ROCm/hip-tests commit: 72e6432739]
Implemented positive and negative test cases for the following APIs:
- hipApiName
- hipKernelNameRef
- hipKernelNameRefByPtr
- hipGetStreamDeviceId
Enabled build of Callback tests only for AMD platforms, as there is no support for Nvidia platforms.
Forgot SDMA blit engine indices are offset by DevToDev 0-position in
a couple of places.
Change-Id: Ie811d8281bc812738ed0107694f3dffde5e93685
[ROCm/ROCR-Runtime commit: 7364a93b98]
The MemoryAllocAll test in kfdtests exercises the new KFD memory
availability API by trying to allocate a single buffer object that
exactly fills all of vram. Desired object size is determined using the
memory availility KFD ioctl via libhsakmt, then an object is allocated
slightly larger than that size. If the allocation attempt fails then
the test tries to allocate a slightly smaller object, and continues
trying with smaller sizes until the allocation succeeds. The test
succeeds if the successfully allocated object is within some specified
tolerance of the available memory reported.
There are a number of known issues that can cause the successfully
allocated object to be significantly smaller than reported availability.
Until these issues are addressed, we should not fail the test, but just
log the actual divergence between the size of the object we thought we
could allocate, and what was actually possible.
Signed-off-by: Daniel Phillips <daniel.phillips@amd.com>
Change-Id: I165a30865ffbb2353286dcc896ad8e24af124615
The MemoryAllocAll test in kfdtests exercises the new KFD memory
availability API by trying to allocate a single buffer object that
exactly fills all of vram. Desired object size is determined using the
memory availility KFD ioctl via libhsakmt, then an object is allocated
slightly larger than that size. If the allocation attempt fails then
the test tries to allocate a slightly smaller object, and continues
trying with smaller sizes until the allocation succeeds. The test
succeeds if the successfully allocated object is within some specified
tolerance of the available memory reported.
There are a number of known issues that can cause the successfully
allocated object to be significantly smaller than reported availability.
Until these issues are addressed, we should not fail the test, but just
log the actual divergence between the size of the object we thought we
could allocate, and what was actually possible.
Signed-off-by: Daniel Phillips <daniel.phillips@amd.com>
Change-Id: I165a30865ffbb2353286dcc896ad8e24af124615
[ROCm/ROCR-Runtime commit: d3bb1ca4af]
Since KFD counts svm allocation as system memory usage,
KFDSVMEvictTest will fail on the case of small system
memory, adding check is to skip test.
Signed-off-by: Eric Huang <jinhuieric.Huang@amd.com>
Change-Id: I040f16f2dd0d4092d069a632cfba9c28293f781b
Since KFD counts svm allocation as system memory usage,
KFDSVMEvictTest will fail on the case of small system
memory, adding check is to skip test.
Signed-off-by: Eric Huang <jinhuieric.Huang@amd.com>
Change-Id: I040f16f2dd0d4092d069a632cfba9c28293f781b
[ROCm/ROCR-Runtime commit: 3f55ba9fb8]
Ioannis Assiouras
Sean Keely
Bill(Shuzhou) Liu
ROCm CI Service Account
Xiaomeng Hou
Marko Veniger
music-dino
nives-vukovic
milos-mozetic
Jonathan Kim
Daniel Phillips
Nirmal Unnikrishnan
Eric Huang
Jatin Chaudhary
Anusha GodavarthySurya
Jenkins
jatang