If cl_khr_fp16 extension is enabled, then OCL runtime should report CL_DEVICE_HALF_FP_CONFIG.
Change-Id: I7c4ac48387f80bc704a475c57e5b52a462090d1b
[ROCm/clr commit: ad2c1e899a]
- Enable CUs adjacent pairwise for WGP mode
- In HostQueue::terminate() do not segfault if virtual device hasn't been created
Change-Id: I94402ff333308af5824878086cc238b3993d534d
[ROCm/clr commit: 8c1232124e]
- Rename HIP_USE_SDMA_QUERY to DEBUG_CLR_USE_SDMA_QUERY as this is
supposed to be a temporary env var for debug purposes only.
Change-Id: If6ebd52ab87624375a3df24ceccdcc05c60a65af
[ROCm/clr commit: 770b2a4711]
- Make IPC interfaces generic between devices and rely on the IPC buffer
for attach/detach logic
Change-Id: Id3c18d122030329b7ee532bbb6317de9dd6a0bbe
[ROCm/clr commit: 9aa6f25f1c]
HIP_FORCE_DEV_KERNARG=1 will create a device allocation for kernel arg
segment. Flag is 0 by default.
Change-Id: Iaaf5a149f3be8596568878d5d272268baf067c60
[ROCm/clr commit: 5436d362b1]
- Maintain a map of SDMA engine# to stream allocated following a greedy
approach
- Anything past that will query SDMA engine status always and go with a
SDMA or Blit copy path
Change-Id: Ibfaed7f951ab84d80cb0430596a4d11b5aec9202
[ROCm/clr commit: 5865c642d4]
This reverts commit dfa7790030.
Reason for revert: Deferred to a future release.
Change-Id: Ia66c37f0ab9734dee73c930d10d7469d5fd57254
[ROCm/clr commit: 5dc104b3ea]
Remove alignment to granularity for IPC handles as ROCr has a patch that
will internally validate pointer sizes against requested size during
allocation instead of size aligned to page size. This patch is needed
together with this patch from ROCr:
f8a42a3a:Use user requested size for memory fragments
Change-Id: I28b25558ea03c836b44fafdb34b7330cf6887424
[ROCm/clr commit: fa1b7925cf]
Heap initialization used device queue, but it shoudl be used for
cooperative launches only. Heap initialization must use the same queue
as the current dispatch.
Change-Id: I856621bf82bbdeb1c2d0fbc4970e90d09af805cb
[ROCm/clr commit: 67af16bdf3]
This reverts commit 7a5c07bf4e.
Reason for revert: performance drop in Quicksilver app
Change-Id: I2bdf42ad0a235a74e2bb4d38f86471e9affbd7a6
[ROCm/clr commit: 3e88081961]
Remove amdgpu-pro interface for persistent memory,
used in Liquidflash
Change-Id: I7d1720ad0875a62ebb2d7f96cba39601d560a5df
[ROCm/clr commit: 6dad2fc306]
This reverts commit 06593a072f.
Reason for revert: ROCr query should now be usable in upcoming release.
Change-Id: I2207761ca6af5d585d090bae1af09eb9a8e9bad6
[ROCm/clr commit: a52f5bda8f]
The ROCclr assigns zero-based IDs to GPUs in the order they are
discovered. That zero-based ID is what is used to identify the GPU
on which the HIP_OPS activity took place.
When multiple ranks are used, each rank's first logical device always
has GPU ID 0, regardless of which physical device is selected with
CUDA_VISIBLE_DEVICES. Because of this, when merging trace files from
multiple ranks, GPU IDs from different processes may overlap.
The long term solution is to use the KFD's gpu_id which is stable
across APIs and processes. Unfortunately the gpu_id is not yet exposed
by the ROCr, so for now use the driver's node id.
Change-Id: Ib78854527d600d175bb76e2df0747c33f898c615
[ROCm/clr commit: 9a82118c85]
- Store last fence scopes and use the last value to determine if we need a cache flush again. This helps cases where hipExtLaunchKernel API is
used.
- Purge code for ROC_EVENT_NO_FLUSH
Change-Id: I531cf9c9c60d5e2b3a9e265d0f52f79ed2fa8a8c
[ROCm/clr commit: 9b5cbd37a2]
- rocr attribute needs to be updated after each iteration
Signed-off-by: sdashmiz <shadi.dashmiz@amd.com>
Change-Id: I3afb2d7954ef3de37f5f5f9d3cc7757fdacffcec
[ROCm/clr commit: 50e0ddb055]
Julia Jiang
Saleel Kudchadker
Alex Xie
German
kjayapra-amd
Rakesh Roy
ajay
Ioannis Assiouras
Jaydeep Patel
Xiaomeng Hou
Anusha GodavarthySurya
Maneesh Gupta
David Yat Sin
jatang
Jiabao Xie
pghafari
Laurent Morichetti
Sourabh Betigeri
sdashmiz
Sarbojit Sarkar