It seems that due to removal of vdev()->isHandlerPending(),
Marker queued to ensure finish is not enqueued and that cause
hung at waiting event for kernel enqueue command.
Change-Id: I364abb2dcb4897b11a7eb61b5d85013b69292792
- Report kernel names for optimized graph path
- Refactor code so that we store profiling info in Accumulate command
Change-Id: Ib97735a0239aeb9fc3a50a4bb7126dd0bcadc8af
For avx build, the start address of values_ buffer in KernelParameters is not
correct as it is computed based on 16-byte alignment.
Change-Id: I3b28ae02d2c9c0517d4a348d95ae8c6721bec83d
- Do not use extra barrier to detect graph end. If its a kernel node we
can use a completion signal for the last packet. Saves roughly 6us for
Phantom testcase per graph launch.
Change-Id: I5e0c2479d9964fbeda86ed97533f6718f49a7f91
- Track all captured commands under a new AccumulateCommand
- Add begin() and end() methods to capture commands
- Explicit TS object now passed to certain methods because
profilingBegin() and profilingEnd() now happen separately and thus can
run into threading issues
Change-Id: I171106bdcad72b057836cb2f3fc398db3533119f
This header file is used by ROCclr and might be used by HIP too.
ROCclr folder is a right place for this header file.
Change-Id: Ie2307331f9bd445f3c295d2ccbfc4924b0fa9510
Some pytorch tests use a tracer plugin and rely on profiling information
to be reported right after hipDeviceSynchronize()
Change-Id: Ib021a1e7b1a30b3c24de72627c471810f7f7878d
Support hipExternalMemoryGetMappedMipmappedArray().
Add ImageExternalBuffer to differiate ImageBuffer.
Currently we only support tiling_optimal mode as
vulkan driver doesn't provide tiling information.
Change-Id: I7e3524cdde53e4df9f728894bcebf4bd3f58d4d9
Use large signal pool if profiler is connected or profiling forced
enabled. This is needed to mitigate signal creation overhead when
profiling as signals are attached to every packet and deeper batch may
show overhead of signal allocation.
Change-Id: I8034b8a20b55328b87d593bf044f59672f9653e8
This reverts commit 44a3935cda.
Implement the right way to make ExternalSemaphores be signalled
only after prior works on the stream have been finished.
Change-Id: I9d5974e05d5f229170b928db4566c14e40e3cbaa
Add a view bit to avoid original resource destruction when parent
dependency doesn't exist with the image view cache
Change-Id: I8277afd575af8f29951c5d1a9f7d94d784251657
Make sure parent_ field is cleared for the internal image views.
The internal image views don't require dependencies tracking.
The issue appeard only when Navi10 pitch workaround was enabled.
Change-Id: I376d212750085a9391f8c32fc2979dcb5d93c89c
- Enable CUs adjacent pairwise for WGP mode
- In HostQueue::terminate() do not segfault if virtual device hasn't been created
Change-Id: I94402ff333308af5824878086cc238b3993d534d
rocclr/platform/external_memory.hpp:93:30: warning: class with destructor marked 'final' cannot be inherited from [-Wfinal-dtor-non-final-class]
93 | virtual ~ExternalImage() final {}
Change-Id: I56d760fa6c08544100e3bc03d35129bd16d8a428
Add missing mipmap Apis’ implementation.
Fix some bugs of mimpmap apis.
Use hipmipmappedArray to differentiate cuda
and driver apis on Nvidia.
Change-Id: I6079d9f3b2ddf4e42b9a6f7f3902322cfca02cfd
- disable deprecated function use warning
- disalbe size_t to .type' warning
- disable conversion from 'type1' to 'type2' warning
Signed-off-by: sdashmiz <shadi.dashmiz@amd.com>
Change-Id: I64161fd37cf56de3d132102103267ae8da40193a
Initial implementation for hipMemPoolExportToShareableHandle,
hipMemPoolImportFromShareableHandle,
hipMemPoolExportPointer and hipMemPoolImportPointer
Change-Id: I0ebdc48e9163b394ded560adca6c38bbc5aee7d1
The change enables VM support in graphs on Windows. That allows
to avoid caching of all allocations at the cost of map/unmap
overhead during memory create/destroy.
Change-Id: I792be00fba099e5e5d3cd44a963e1dfd6976a86d
- The implementation in mempool graphs requires refcounting VA object.
That requires release() to update the map only on the actual destruction.
- Add GPU event tracking for paging operation. Otherwise, runtime
may not always flush IB.
Change-Id: Idf99ffb894321a38e04b490116a7ca435635918d
For large bar platform, it's not necessary to mmap
memory allocated on VRAM to cpu again
Change-Id: I0701680476829d4058b3e7b643e8df657d0c6168
Signed-off-by: Ruili Ji <ruiliji2@amd.com>
OS can terminate unfinished queue thread from default stream at any
time. Potentially leaving the queue lock in a bad state and causing a
deadlock if runtime destroys VirtualGPU later from the host thread.
Change-Id: I247f102ee84e6b4dba947504933395071945c85d
Windows kills threads on exit without any notification. However,
runtime can still destroy VirtualGPU object from the host thread with
HostQueue destruction.
This change also forces RGP trace transfer on the last capture without
any delays.
Change-Id: I768e87e99e1d23a021e63c12f36e450817743759
Jaydeep Patel
Saleel Kudchadker
Ioannis Assiouras
kjayapra-amd
jiabaxie
German Andryeyev
Alex Xie
taosang2
Anusha GodavarthySurya
Ajay
Todd tiantuo Li
Rakesh Roy
Matt Arsenault
sdashmiz
victzhan
Ruili Ji