Extension allows to execute the kernels without a wait barrier and L1
invalidation.
Change-Id: I96c485204303f54a0240b93134f4560673e4bd17
[ROCm/clr commit: 13c6f56ca9]
Dont track the status of fence_dirty_ flag on the host, instead clear it
when we submit a barrier on the respective stream.
Change-Id: I4d98dbf20c81379c9c5da9f5b67629a8f9f6dfcd
[ROCm/clr commit: 0b0df605d4]
- Add API table versioning defines to hip_api_trace.hpp
- Add rocprofiler-register implementation in hip_api_trace.cpp
- created ToolInit implementation
- moved some local functions into anonymous namespace
- renamed some function to be overloads
- made most of the initialization code templated (reduced code duplication)
- enforce ABI
- Updated hipamd/src/CMakeLists.txt
- find rocprofiler-register package (enabled by default but not required)
- Address review comments
- simplify size calculation for dispatch table
- remove setting .size field in GetDispatchTableImpl
- Fix windows build
- Do not enforce ABI on Windows
Change-Id: I08766e8a083528a1236996274bf4522e0e8a1b9f
[ROCm/clr commit: 0f1fdc7794]
Use only 16 workgroups for compute P2P copies.
That should be enough to utilize XGMI bandwidth.
Change-Id: I60dfe019279bb95f93c8874244c1738aad1896d8
[ROCm/clr commit: 31101c6219]
The following builtins from the CUDA spec are implemented:
- __all_sync, __any_sync, __ballot_sync and __activemask
- __match_any_sync and __match_all_sync
- __shfl_sync, __shfl_up_sync, __shfl_down_sync, and __shfl_xor_sync
The following builtins are NOT implemented, pending support in the compiler:
- __reduce_add_sync, __reduce_min_sync, __reduce_max_sync
- __reduce_and_sync, __reduce_or_sync, __reduce_xor_sync
Change-Id: I07dedbbfe5449f4b5c9b040bed59f5603ccec8c3
[ROCm/clr commit: c5ab5680b4]
- Use isFenceDirty() check in hipStreamQuery.
In case if there is a stopEvent for hipExtLaunch* API we avoid system
scope if there are consecutive hipExtLaunch* and replace them with agent
scope, but then if one uses hipStreamQuery it should detect we need a
barrier with system scope.
Change-Id: I539810f659d4e1c49104e3a0da6ff2803c6b3dee
[ROCm/clr commit: 4b3c4709ff]
This reverts commit 4abdfe5489.
Reason for revert: multiple tickets filed. i will re-instate this change after fixing those.
Change-Id: I66152e3643c73ec1d240d0fcddf2119f8f4848a8
[ROCm/clr commit: 5f7ebda6aa]
This issue is due to the fact that we set fence_dirty_ flag for every
kernel launch. Whenever any next HIP API is caused, the stream logic
assumes fence is dirty and queues a marker, when its not needed.
Change-Id: I10a49ee951daa92221aa6be75e2334849579a45d
[ROCm/clr commit: d6b4892600]
Recently some unused compiler options for HSAIL path were removed,
but it affected blit kernels compilation. Hence, remove those options.
Also delete assert for device to device copy in SDMA path for now.
Change-Id: Ib5d7f063af2ab4a3fc5d73d426e39c391b1011ac
[ROCm/clr commit: eaa61fc740]
Luxmark still uses HSAIL path and one subtest can benefit from the wave limit.
Change-Id: I16c94e09cd6e2afd6341cb76bf2e9ab7b7713214
[ROCm/clr commit: dec1158d04]
Single OCL context with multiple devices should use direct map
even with persistent memory, because cache coherency layer.
Change-Id: I02a48830440203a1dea176acd4dff4452b70f133
[ROCm/clr commit: 49037b9762]
- Make sure persistent memory from resource cache is properly adjusted
in free memory calculation.
Change-Id: I74ef68975ccde4694fb1cb904617c418e85dfc9f
[ROCm/clr commit: 85c15d720d]
hipIpcOpenMemHandle can be used to open memory handle and fetch
a device pointer on a different device than the actual device
where the memory was allocated. The device Id must be set to
the current device in such cases and not the original device.
Change-Id: Ie1c7eada928d02124a41125876876f96015552e5
[ROCm/clr commit: c398c75512]
This is an initial change before we refactor the build/link paths for
kernel launches for HIP. This current change is needed as compiler was
setting some dump file which needed fs access which has slowdowns for
NFS mounted file systems
Change-Id: I828f9bb04d789b4f8c05c1ed08767f325efeb47c
[ROCm/clr commit: 3f2f7252aa]
The function erroneously returns hipErrorInvalidValue when kind
is set to hipMemcpyHostToHost
Change-Id: I6810b9f04f3218e517fd2f96410d1375e6ac6ff0
[ROCm/clr commit: af86c1b8a2]
Sorted map can work much faster for many allocations and a low reuse frequency
Change-Id: I6dba29ebc8bfacdf34307149b6a2b194890b2932
[ROCm/clr commit: a1fffac595]
Add GPU_DEBUG_ENABLE to control ttpm behavior. If enabled,
then HW will collect more debug info at some perf cost
Change-Id: Icee0686b903a7b1bd483710b9d611877cd43c6aa
[ROCm/clr commit: 7d661bc7df]
- For new AccumulateCommand we enqueue nop barrier packet. So during stream sync we need system flush.
Change-Id: I0b97b626bcdae582ef95e7c95030d78df1fa5a54
[ROCm/clr commit: 748c6b3520]
Runtime has to release extra memory, held by the pools,
in synchronization points for event, stream or device.
Change-Id: Id533a5e1d137812aa72bdfe101b4b333c6a43d66
[ROCm/clr commit: 3fa4e31180]
When an offset is applied to the source or destination pointers plus the kind is set to
hipMemcpyDefault and the source or destination is allocated with hipMallocManaged
hipMemCpy2D erroneously fails with hipErrorInvalidValue.
Change-Id: I0db4c17514f743652d8f9a2691da6601a2abb2a1
[ROCm/clr commit: d3bfb55d7a]
Fix sporatic segment fault in texture test
via retaining image in texture object which
references the image.
The image will be released when the texture
object is destroyed.
Change-Id: Ic3fefa2d5dda6afebd1acd4d41ad310b138af6dd
[ROCm/clr commit: d6d235a111]
pghafari
German
kjayapra-amd
Maneesh Gupta
Saleel Kudchadker
Jonathan R. Madsen
sdashmiz
Jatin Chaudhary
Sameer Sahasrabuddhe
jujiang
Konstantin Zhuravlyov
David Salinas
Rahul Garg
ksankisa
Satyanvesh Dittakavi
Ioannis Assiouras
Konstantin Zhuravlyov
Jaydeep Patel
Anusha GodavarthySurya
Alex Xie
taosang2