Setting AMD_CPU_AFFINITY = 1 will make runtime honor core affinity that
the process may set. This is disabled by default as it can prevent
worker thread or any thread that runtime creates from getting scheduled
thus affecting performance.
Change-Id: Ibe4cc95e7b99caee5ce750b7bf66e09e999cc9a3
[ROCm/clr commit: 1398719b0d]
HIP should be built with HSAIL support disabled.
Currently HSAILProgram::info() and VirtualGPU::buildKernelInfo() expose
ACL interfaces directly. This should not be allowed.
Change-Id: Iae15d4f19be16806826f2f6cb600752c11f97fc1
[ROCm/clr commit: bbe6246f19]
hip_rtc.cpp refers to the symbol __hipGetPCH(), but that one can only be
found in hip_global.cpp. It's not referenced anywhere else, hence it
should be moved.
Change-Id: Ifc67b2f3b1b8344dec0e6c30d083d62288fb5e24
[ROCm/clr commit: 20eff6c298]
hipGraphAddHostNode and hipGraphAddMemcpyNode1D are not part of the HIP
graph implementation, hence they should not be exported.
Change-Id: Ia3d0188895fb4f64ebfa23d4a0a09c44438f878d
[ROCm/clr commit: 0782883cb3]
- Accept hipMemAttachHost flag
- Don't allocate HMM memory if HMM is disabled in KFD
Change-Id: I3d386eb0a109a19c16a87a728944a3a6669f643a
[ROCm/clr commit: 5d8cafef2d]
For hipRTC on Windows, add macro __HIPCC_RTC__ to allow
online compilation of with device functions excluding standard
C/C++ headers, system headers, and host HIP APIs.
Change-Id: I1d91f042baf1359856ec83ab7030dc58785e0334
[ROCm/clr commit: 688693e9fd]
On StreamBegincapture captures the parameters passed to APIs and respective node will be created and added to graph
All parameters are passed to STREAM_CAPTURE macro, it checks if stream in capture mode and redirects the call to the capture function and returns
Updated hipStream and hipEvent with capture parameters
Added handling for hipStreamBeginCapture & hipStreamEndCapture
Change-Id: Ic8926a7b4336c2cc81f0b3a9a224aa392c474134
[ROCm/clr commit: 8cc0e04239]
Currently LiquidFlash cannot be supported from Github Enterprise,
hence we need to be able to build with out it.
Allow this by setting -DWITH_LIQUID_FLASH=0.
Change-Id: If59afe45c02a1623639360dbcc3927360f16ebd3
[ROCm/clr commit: ff3f6b2575]
Currently LiquidFlash cannot be supported from Github Enterprise,
hence we need to be able to build with out it.
Allow this by setting -DWITH_LIQUID_FLASH=0.
Change-Id: I975e8ee16b7ba033e3eb95fe40955d8c1d4779b7
[ROCm/clr commit: 7034e749e3]
Currently LiquidFlash cannot be supported from Github Enterprise,
hence we need to be able to build with out it.
Allow this by setting -DWITH_LIQUID_FLASH=0.
Change-Id: I8f147b5d52ca1c8203000beea8f16b2699548dc3
[ROCm/clr commit: 8af3f12790]
runtime directory only exists in P4. We should not be referencing it.
Change-Id: I1d88f32fccc716468c070235c4797bd905a76866
[ROCm/clr commit: 5958b3bba5]
aclutGetTargetInfo() is an internal compiler lib helper functions. This
will not be imported in the HSAIL shared library build, however it is
simple enough that we can maintain our own local copy of it.
Change-Id: I91d1a336c7da027bf8a7df8fae86a25add533611
[ROCm/clr commit: 7fd1e9c10a]
Selector indices are as follows (the upper 16-bits of the selector are not used): selector[0] = s<2:0> selector[1] = s<6:4> selector[2] = s<10:8> selector[3] = s<14:12>
Change-Id: Ibf76c6ec2374f1f5b9bba8bd9dbd73660f830eea
[ROCm/clr commit: e9cc056d25]
Windows doesn't have headers unistd.h, sys/wait.h, so
remove them for _WIN64. Also, Windows doesn't have rand_r
so define it. Fixes builds for HIP tests:
d2dMemCpyWithPinnedHostMemory, hipMemset3DFunctional,
hipModuleLoadMultProcessOnMultGPU,
hipModuleLoadUnloadStress, hipModuleNegative,
hipStreamACb_ThrdBehaviour, hipMemcpyNegativeMThrdMSize,
hipStreamACb_MStrm_Mgpu, StreamAddCallback,
hipStreamACb_StrmSyncTiming, hipStreamACb_MultiCalls,
hipMallocConcurrency, hipMemset3DRegressMultiThread,
and hipSetGetDevice.
Change-Id: Id2f98d664716026c590d879e464ba467d4c071be
[ROCm/clr commit: bacb89c086]
Fix wrong mixing of current device and stream device in
ihipModuleLaunchKernel() and hipLaunchCooperativeKernel().
Fix missing hipSetDevice() in hipMemcpyWithStream* tests.
Change-Id: I09333bb40d239bb42c832df5ea16d17eeaeff5e7
[ROCm/clr commit: b99efb9d67]
Change-Id: I5daeacd9dd5c6ce7f914d6e6e45dd41fb2a675a5
hipMemRangeGetAttributes was returning hipErrorInvalidValue due to improper
mapping of the arguments to cudaMemRangeGetAttributes.
[ROCm/clr commit: cc39ceab25]
Jason Tang
Chauncey Hui
German Andryeyev
Saleel Kudchadker
Sourabh Betigeri
Vladislav Sytchenko
Julia Jiang
pghafari
Brian Sumner
Aaron En Ye Shi
Anusha Godavarthy Surya
Ravi C Akkenapally
Maneesh Gupta
Christophe Paquot
kjayapra-amd
Jatin Chaudhary
Satyanvesh Dittakavi
Tao Sang
Anton Gorenko